Bearing Fault Frequency Calculator (BPFO, BPFI, BSF, FTF)

SteelSolver.com · Vibration & Reliability

Bearing Fault Frequency Calculator — BPFO, BPFI, BSF & FTF

Enter the bearing geometry and shaft speed to get all four characteristic defect frequencies in Hz, CPM and shaft orders, with harmonics, sidebands, a one-sided slip tolerance band, an editable bearing cross-section and a peak-matching tool you can take straight to the spectrum analyser.

SteelSolver.com · Vibration & Reliability

Bearing Fault Frequency Calculator — User Guide, Field-by-Field Walkthrough & Engineering Notes

This guide explains every input field on the Bearing Fault Frequency Calculator, shows the units and expected value ranges, and describes how the four characteristic defect frequencies — BPFO, BPFI, BSF and FTF — are derived from bearing geometry and shaft speed. It also covers accuracy expectations, common data-entry mistakes, and worked examples you can reproduce by hand. Use it as a training reference or as a quick field guide when you are standing at the analyser with a bearing datasheet in hand.

What Is This Bearing Fault Frequency Calculation Used For?

The calculator converts a bearing's internal geometry into the four characteristic defect frequencies that vibration analysts use as cursor positions on a spectrum. Instead of guessing where a bearing defect should appear, you get an exact, speed-dependent value in Hz, CPM and shaft orders, together with the harmonic family and sideband pattern that confirm a genuine fault.

In practical reliability work the calculation is used for:

• Baseline setup — pre-loading defect frequencies into a vibration database so each measurement is automatically checked against the correct lines.
• Diagnosis — deciding whether a peak at, say, 162 Hz is an inner-race defect or an unrelated 5.4× running-speed harmonic.
• Trending — tracking the growth of a defect family over weeks, long before it becomes audible or shows in a velocity spectrum.
• Root-cause investigation — ruling a bearing in or out when a machine shows elevated vibration but no obvious synchronous pattern.
• Spare-parts planning — quantifying how much warning time the defect frequencies give before a planned outage.

Step-by-Step Guide — Filling In Every Field

The calculator is organised as a single input form on the left (or top, on mobile) and a live result panel on the right. Every time you change a value the four frequency cards, the summary table, the calculation steps, the diagram and the spectrum map update immediately. Below is a field-by-field walkthrough.

1. Bearing Preset

   What it does: selects a pre-loaded set of geometry values for a well-known bearing, so you do not have to type the numbers yourself. The three presets cover:

   • 6205 — deep groove ball, Z = 9, d = 7.938 mm, D = 38.50 mm, φ = 0°.
   • 6312 — deep groove ball, Z = 8, d = 22.225 mm, D = 95.00 mm, φ = 0°.
   • 6316 — deep groove ball, Z = 8, d = 28.575 mm, D = 125.00 mm, φ = 0°.

   Units: the preset values are always in millimetres and degrees, regardless of the length unit selector further down.

   Common mistake: assuming a preset matches your bearing just because the bore size is right. A 6205 from one manufacturer can quote D = 38.50 mm and another D = 39.04 mm — a small difference that moves BPFO from 3.572× to 3.594×. Verify against the datasheet of the bearing you actually have in the machine.

2. Input Method

   What it does: tells the calculator which data you actually have, so it can choose the right set of equations.

   • Geometry (Z, Dw, Dp, α) — the full set of classical equations. This is the most accurate mode and the one to use whenever a datasheet is available.
   • Published frequency factors — enter the four orders (× f_r) directly from a manufacturer's catalogue or a vibration database. Useful when you cannot get the raw geometry.
   • Element-count estimate — uses the rough approximations BPFO ≈ 0.4 Z f_r and BPFI ≈ 0.6 Z f_r. Good to a few percent when d/D is near 0.2. BSF and FTF cannot be estimated this way.

   Common mistake: using the element-count estimate to sign off a report. The estimate is only for placing a cursor at the start of an investigation — always confirm with real geometry or with a published factor before making a maintenance decision.

3. Rolling Element Type

   What it does: distinguishes a ball bearing from a roller bearing so the calculator can add a contextual warning about how each behaves in a real spectrum.

   • Ball — point contact, sharp defect lines, strong BSF and 2×BSF energy.
   • Roller (cylindrical / spherical) — line contact, the defect energy is usually smeared into a broad band rather than a single spike, and the BSF line is often missing.

   Common mistake: reporting a missing BSF line as "no defect" on a roller bearing. Roller bearings do not show a clean element-spin line the way ball bearings do; look for 2×BSF with FTF sidebands instead.

4. Rotating Race

   What it does: tells the calculator which race is stationary and which is rotating. In a normal shaft application the inner ring rotates with the shaft and the outer ring is fixed in the housing — that is the default.

   • Inner ring rotates (normal shaft) — BPFO is the lower of the two pass frequencies, BPFI is the higher.
   • Outer ring rotates (wheel hub, pulley) — the roles swap, so the inner-race line becomes the higher one and the outer-race line becomes the lower one.

   Common mistake: leaving the setting on "inner ring" for a wheel-hub or pulley bearing, then searching for BPFO at the wrong value. When the outer ring rotates, the two pass frequencies swap positions in the spectrum.

5. Shaft Speed

   What it does: sets the fundamental frequency f_r from which every defect frequency is scaled. All four characteristic frequencies are directly proportional to shaft speed.

   Units: RPM or Hz, selected with the unit dropdown to the right of the number. Internally the calculator converts RPM to Hz using f_r = RPM / 60, and reports both Hz and CPM in the results.

   Common mistake: using the nameplate RPM when the machine is on a VFD or a belt drive. Read the actual 1× peak from the spectrum and enter that value — it is the measured running speed that determines where the defect lines actually sit.

6. Slip Tolerance Band

   What it does: defines the width of the expected measured band around each theoretical frequency, expressed as a percentage below the calculated value.

   Units: percent (%). The default is 2 %. The upper side of the band is set to one quarter of the lower side, because real peaks drift low under slip and rarely drift high.

   Common mistake: setting the tolerance to zero and then rejecting a genuine defect peak because it sits 1.5 % below the theoretical value. The classical equations assume pure rolling with no slip; real bearings skid.

7. Bearing Geometry — Z, d, D, α

   These four fields are the heart of the geometry method. Each one has a dedicated label and unit hint on the calculator.

   Number of Rolling Elements (Z)

   What it is: the total count of balls or rollers in the bearing. Units: a dimensionless integer, normally between 6 and 20 for common bearings. Effect: Z drives BPFO and BPFI linearly — an error of one in Z shifts both pass frequencies by roughly 10 %.

   Common mistake: counting the visible balls on one side and forgetting the ones behind the cage. Recount from the datasheet.

   Rolling Element Diameter (d)

   What it is: the diameter of a single ball or roller. Units: millimetres or inches, selected with the unit dropdown to the right of the number. Only the ratio d/D matters, so any consistent unit works.

   Common mistake: measuring the raceway width instead of the element itself. Use the datasheet value; a caliper on a used bearing is rarely accurate enough.

   Pitch Diameter (D)

   What it is: the diameter of the circle that passes through the centre of every rolling element — roughly midway between the bore and the outside diameter. Units: the same length unit as d.

   Common mistake: entering the bore diameter or the outside diameter. D is not the bore and not the OD; it is the centre-to-centre circle through the elements.

   Contact Angle (α or φ)

   What it is: the angle between the line of action through the rolling element and the radial plane. Units: degrees. Typical values: 0° for deep groove ball bearings, 15–40° for angular contact, up to 45° for some self-aligning designs.

   Common mistake: leaving the value at 0° for an angular contact bearing. Ignoring a 25° contact angle shifts every frequency by several percent — enough to miss the defect entirely.

8. Published Frequency Factors (Factor Mode Only)

   When the input method is set to "Published frequency factors", four number boxes appear instead of the geometry fields. Each expects an order value (a multiple of shaft speed) taken from a bearing catalogue or a vibration database.

   Units: orders (× f_r), dimensionless. Typical values: BPFO around 3.5–4.5, BPFI around 5–7, BSF around 2–3, FTF around 0.38–0.42.

   Common mistake: swapping BPFO and BPFI factors. BPFI is always the higher of the two. The calculator warns you if the entered BPFI is lower than BPFO.

9. Harmonics to List

   What it does: controls how many integer multiples of each fundamental frequency are shown in the harmonics table and used by the peak-matching tool.

   Units: none — it is a count from 1 to 10. The default is 1–3×, which is the usual range for a developing defect. Increase it to 5× or 10× when you are dealing with a well-established fault that shows many harmonics.

10. Action Buttons

    Five buttons sit below the input fields. Each performs one specific job:

    • Copy result summary — copies a plain-text report (bearing name, speed, geometry, all four frequencies with CPM and orders) to the clipboard, ready to paste into a CMMS or an email.
    • Download CSV — saves a CSV file containing the fundamentals, the expected band and the harmonic series, suitable for import into Excel or a vibration database.
    • Print / save PDF — opens the browser print dialog with a print-optimised layout that includes the input values and all result tables.
    • Copy share link — encodes the current inputs in the URL so you can send the exact setup to a colleague.
    • Reset — returns every field to the default 6205 values.

Worked Example — 6205 Bearing at 1,800 RPM

This example uses the default preset and can be reproduced by hand to check your understanding of the equations. Geometry: Z = 9, d = 7.938 mm, D = 38.50 mm, φ = 0°. Shaft speed 1,800 RPM, so f_r = 30 Hz.

Sanity check

Two identities must always hold for a correctly calculated bearing: BPFO + BPFI = Z × f_r, and BPFO = Z × FTF. Here: 107.2 + 162.8 = 270.0 Hz, and 9 × 30 = 270.0 Hz. Also, 9 × 11.91 = 107.2 Hz. Both identities hold, so the arithmetic is sound. The calculator prints this check automatically beneath the result cards.

Full result table

Remember that a single ball defect strikes both races once per element revolution, so the dominant measured line for a ball defect is usually 2×BSF (139.3 Hz here), often flanked by sidebands spaced at FTF.

Accuracy Note — What These Numbers Do and Do Not Guarantee

The four equations implemented by this calculator are ideal kinematic formulas. They assume pure rolling with no slip, rigid rings, a perfectly rigid cage and a constant shaft speed. Real bearings depart from those assumptions in three important ways:

• Slip. Under light load, grease drag or high speed the element set skids, so the measured peak typically falls 1–2 % below the theoretical value. That is why the calculator shows a one-sided expected band rather than a single dotted line.
• Pitch diameter tolerance. Different catalogues quote slightly different pitch diameters for the same bearing number. The resulting spread is usually smaller than the slip effect, but it explains why two databases can disagree in the third significant figure.
• Modulation. A defect that rolls through the load zone is amplitude-modulated, producing sidebands at f_r (around BPFI) or at FTF (around BSF). A single peak with no family or sidebands is weaker evidence than a clean harmonic series.

Use the calculator to decide where to look. Confirm a diagnosis with the envelope spectrum, with the harmonic family and with the sideband pattern. A peak that sits 5 % above the calculated frequency is almost never the bearing.

Common Input Mistakes and How to Avoid Them

Key User Pain Points and How This Calculator Solves Them

Where Engineers Apply This Calculation

• Condition monitoring programmes — every route-based vibration survey starts by loading the correct defect frequencies into the analyser, so each measurement is automatically checked against the right lines.
• Reliability engineering — calculating how far a defect family has developed tells you whether the bearing will survive to the next planned outage or needs an emergency work order.
• Failure analysis — after a bearing has failed, the calculated frequencies confirm whether the damage pattern seen on the raceway is consistent with the vibration evidence that preceded it.
• Machine acceptance testing — a new or rebuilt machine can be screened for bearing defects before it enters service by checking its spectrum against the calculated families.
• Spare-parts and maintenance planning — matching the predicted defect frequency to the observed trend gives the planning team a defensible replacement window.

Real-World Usage Example — Motor Drive-End Bearing

A 75 kW induction motor drives a centrifugal pump through a belt. The vibration analyst sees a rising peak at about 160 Hz in the velocity spectrum and an elevated noise floor in the 2–8 kHz acceleration band. The nameplate speed is 1,485 RPM, but the measured 1× peak is at 24.5 Hz (1,470 RPM) because the belt slips slightly.

1. The analyst identifies the drive-end bearing as a 6312 and selects the preset. Shaft speed is entered as 24.5 Hz using the measured 1× peak, not the nameplate 1,485 RPM.
2. The calculator returns BPFI = 133.1 Hz (5.428×). The observed 160 Hz peak is well above that and does not match. The analyst checks the 6312 harmonics: 2×BPFI = 266.2 Hz, 3×BPFI = 399.3 Hz. Still no match.
3. The analyst widens the search. The peak at 160 Hz is actually 6.53× running speed — a synchronous harmonic, not a bearing defect. The real issue is belt-drive misalignment.
4. The bearing is ruled out, and the maintenance team avoids an unnecessary motor strip-down. The calculated bearing frequencies did their most valuable job: they proved the bearing was not the problem.

Frequently Asked Questions

SteelSolver tools are for engineering guidance. Results depend on the accuracy of the geometry you enter and on the assumptions listed above; always verify against the bearing manufacturer's own data before acting on a maintenance decision.