Plain Bearing PV Limit Calculator

PV = p·v on the PROJECTED area for a dry or boundary-lubricated bush, with sourced PV, pressure and velocity limits for eleven materials — and the minimum bush length the longest of the three demands.

Last updated 24 Sep 2026

Plain bearing PV limit

Load, diameter, speed, material → minimum length

Radial load W (N)

Shaft diameter d (mm)Also the bush bore. The pressure uses the PROJECTED area d·L, not the wrapped area πdL — projected, and the difference is a factor of π in the flattering direction.

Bush length L (mm)What you have, or what you propose. The headline answer is the minimum this duty needs; compare the two.

Rotational speed (rev/min)Set it to zero for a static or oscillating duty, where PV does not apply and the pressure limit alone governs. Note that an oscillating bush has its own problem — the film never establishes and the wear is concentrated at the reversal.

Bush materialPV limits are published in psi·ft/min in the US sources these come from; the metric equivalent is shown beside each. One psi·ft/min is 3.5025 × 10⁻⁵ N/mm²·m/s exactly, and mixing the two units is the commonest PV mistake.

DutyThe published PV limits are maxima at a stated duty. QBC’s sintered bronze page is the one source this batch found that quantifies the derate: 20,000 against a maximum of 50,000 for “long-time running with no additional lubrication”, a factor of 2.5. Applying the same 0.4 to the other materials is this site’s extrapolation, not the publishers’.

Not a circuit: the PV limit as the hyperbola it is, and the same hyperbola for every material on the list. That works because both axes are normalised — the vertical one to the material's own pressure limit and the horizontal one to the velocity at which the PV limit and the pressure limit cross, which is PV/p_max. In those units the hyperbola is exactly y = 1/x and the flat pressure limit is exactly y = 1, so they always meet at (1, 1) whatever material you choose. The hatched region under both lines is where the duty has to sit, and the crosshairs mark where yours actually is. Read which side of the crossing you are on: to the left the flat pressure limit governs and the speed is irrelevant, to the right the hyperbola governs and every increase in speed costs pressure capacity. A duty far to the right — a fast, lightly loaded bush — is a heat problem and wants a longer bearing or a better heat path; one far to the left is a strength problem and wants a stronger material.

10.76mmExample

A 1,200 N radial load on a 20 mm shaft at 300 rev/min, in an oil-impregnated sintered bronze bush, intermittent duty

Three limits, and the longest bush any of them demands

p = W/(d·L) · v = πdn/60000 · PV = p·v · L ≥ W/(d·p_max) · L ≥ W·v/(d·PV_max) · v ≤ v_max

p: pressure on the PROJECTED area, d×L — the shadow the bush casts, not its curved surface. Using the wrapped area understates it by a factor of π
v: sliding velocity at the bore, in m/s. Note that the published limits are usually in ft/min
PV: the product, and a proxy for the heat generated per unit of bearing area. It is not a stress and the same PV at high load and low speed is a different problem from low load and high speed
p_max: the material’s pressure limit. For a sintered metal, the DYNAMIC value — four times lower than the static one for bronze
v_max: the material’s velocity limit, independent of the load. A pass/fail rather than a length requirement
L/d: 0.5 to 1.0 on one published band and 0.3 to 1.0 on another. Above about 1 the shaft’s slope governs rather than the pressure

Worked example

A 1,200 N radial load on a 20 mm shaft at 300 rev/min, in an oil-impregnated sintered bronze bush, intermittent duty
Sliding velocity first: v = πdn/60000 = π × 20 × 300/60,000 = 0.3142 m/s, which is 62 ft/min. Sintered bronze's velocity limit is 1,200 ft/min (6.10 m/s), so that hurdle is cleared with a great deal to spare and the velocity limit will not be what governs
The material's limits, in metric units. The published PV limit is 50,000 psi·ft/min, and one psi·ft/min is exactly 3.5025 × 10⁻⁵ N/mm²·m/s, so that is 1.751 N/mm²·m/s. The DYNAMIC pressure limit is 2,000 psi = 13.79 N/mm² — note dynamic, because the static limit for the same material is four times higher and it is the static figure that gets quoted
NOW THE TWO LENGTHS, and the larger of them is the answer. The pressure limit needs L ≥ W/(d·p_max) = 1,200/(20 × 13.79) = 4.35 mm. The PV limit needs L ≥ W·v/(d·PV_max) = 1,200 × 0.3142/(20 × 1.751) = 10.76 mm. So the PV limit governs, by a factor of 2.5, and the minimum bush length is 10.76 mm
Check it the other way round with the bush you have. At L = 20 mm the projected-area pressure is W/(d·L) = 1,200/(20 × 20) = 3.0 N/mm² — PROJECTED, meaning d × L = 400 mm², not the wrapped πdL = 1,257 mm². Use the wrapped area and you get 0.955 N/mm², understating the pressure by a factor of π and the PV by the same
PV at the duty is 3.0 × 0.3142 = 0.9425 N/mm²·m/s, against the limit of 1.751 — 54 per cent of it. The pressure on its own is only 22 per cent of the pressure limit and the velocity is 5 per cent of the velocity limit, which is why it is the PRODUCT that decides this duty and why a bush sized on pressure alone would have been less than half long enough
Two qualifications on the limit itself, both important. It is quoted at a stated duty: QBC's own sintered bronze page says that “for long-time running with no additional lubrication, 20000 should be a limit”, against its stated maximum of 50,000 — a factor of 2.5. Switch the duty selector to continuous and the required length becomes 26.9 mm, well over what you have. And the L/d ratio: at 10.8 mm on a 20 mm shaft that is 0.54, comfortably inside the published 0.5 to 1.0 band
What this calculation is NOT. PV is a heat-generation proxy for a dry or boundary-lubricated bush. It says nothing about a HYDRODYNAMIC bearing, where a pressurised oil film separates the surfaces and the governing quantities are the Sommerfeld number and the minimum film thickness. That is an ISO 7902 calculation, it is a different page, and this one refuses it rather than pretending a PV limit covers it

PV limits, pressure limits and velocity limits by material

Three limits, not one, and the smallest of them governs. Note first the unit trap: these values are published in psi·ft/min in the US sources they come from, and one psi·ft/min is exactly 3.5025 × 10⁻⁵ N/mm²·m/s (built from the exact definitions of the pound-force, the inch and the foot). A PV limit of 50,000 is 1.75 in metric units, and quoting the wrong one is off by a factor of nearly thirty thousand. Note second that the pressure column for the sintered materials is the DYNAMIC limit, which is what a rotating duty needs — it is four times lower than the static limit for bronze (2,000 psi against 8,000), and the static figure is the one most often quoted. Note third that every one of these is quoted at a stated duty: QBC’s sintered bronze page is the only source this batch found that quantifies the derate, saying that “for long-time running with no additional lubrication, 20000 should be a limit” against its stated maximum of 50,000. The last column applies that same factor of 0.4 to every row, which is this site’s extrapolation from one published statement and is labelled as such. These dimensions come from a published standard’s table, not from a formula. The standard itself is cited below and the printed values are attributed to the catalogue they were taken from; a different publisher may round differently in the last digit. A drive is rated for the duty it sees, not for the power it nominally transmits. The service factor used here is stated; the manufacturer’s own factor for your machine class and daily running hours takes precedence over any general table.
Material	PV (psi·ft/min)	PV (N/mm²·m/s)	p max (psi / N/mm²)	v max (ft/min / m/s)	T max	PV derated for continuous duty (N/mm²·m/s)
Sintered bronze, oil impregnated (SAE 841)	50,000	1.751	2,000 / 13.79	1,200 / 6.10	not given	0.701
Sintered lead-bronze	60,000	2.102	800 / 5.52	1,500 / 7.62	not given	0.841
Sintered iron	30,000	1.051	3,000 / 20.68	400 / 2.03	not given	0.420
Sintered copper-iron	35,000	1.226	4,000 / 27.58	225 / 1.14	not given	0.490
Nylon, dry	3,000	0.105	2,000 / 13.79	600 / 3.05	200 °F / 93 °C	0.042
Acetal, dry	3,000	0.105	2,000 / 13.79	600 / 3.05	200 °F / 93 °C	0.042
PTFE, unfilled	1,000	0.035	500 / 3.45	50 / 0.25	500 °F / 260 °C	0.014
Filled PTFE	10,000	0.350	2,500 / 17.24	1,000 / 5.08	500 °F / 260 °C	0.140
PTFE fabric / woven liner	25,000	0.876	60,000 / 413.69	150 / 0.76	500 °F / 260 °C	0.350
Carbon-graphite	15,000	0.525	600 / 4.14	2,500 / 12.70	750 °F / 399 °C	0.210
Phenolic laminate	15,000	0.525	6,000 / 41.37	2,500 / 12.70	200 °F / 93 °C	0.210

Projected, not wrapped — the mistake that flatters by a factor of π

The projected area is the bore diameter times the length — the shadow the bush casts, not the surface it has. Using the wrapped area understates the pressure by a factor of π, which is a factor of π on the PV as well, and it always understates rather than overstates. The reason projected area is the right choice is that the load is not carried uniformly around the bore: it is carried by the arc that the shaft is pressed into, with the pressure falling to zero at the sides, and the integral of that distribution over the loaded arc is exactly the load divided by the projected area. It is not a convention, it is the mean pressure over the load-carrying arc. Every published PV limit is quoted on the projected area.
	Area used	For this page’s example	Pressure it gives
CORRECT: projected area	d × L	20 × 20 = 400 mm²	3.000 N/mm²
WRONG: wrapped area of a full bush	π × d × L	1,257 mm²	0.955 N/mm²
WRONG: wrapped area of a half bush	π/2 × d × L	628 mm²	1.910 N/mm²

Length-to-diameter ratio: the published bands, and they disagree

Two published bands and they do not agree, so both are printed. What they agree on is the top of the range: somewhere around 1.0 the thing that governs stops being the projected-area pressure and becomes the load DISTRIBUTION along the bush — because the shaft slopes, and a sloping shaft in a long bush carries its whole load on one edge. That makes the shaft’s own slope at the bearing the number that decides a long bush, not its PV. Below about 0.5 the bush stops supporting the shaft against angular movement at all and the shaft has to be located somewhere else. Note that the hydrodynamic optimum near 0.4 is a different criterion from the dry-bush band and should not be read across.
Source	Band	Reason given
Armoloy, on plain bush (sleeve/journal) bearings	L/d about 0.5 to 1.0	“Shorter aids misalignment tolerance”. A long bush relies on the shaft staying parallel to it, and a shaft that slopes puts the whole load on one edge
MITcalc, on hydrodynamic plain journal bearings	L/d 0.3 to 1.0, with the load capacity peaking near 0.4	“The bearing has the highest load capacity at B/D ≈ 0.4”. A short bearing loses pressure out of the ends, a long one needs more oil flow to keep cool
Neither	Nothing supporting a band reaching 1.5	This batch looked for a published 0.5 to 1.5 band and did not find one. Above L/d ≈ 1 both sources shift the governing consideration from pressure to edge loading and heat removal

Three limits, the projected area, and the boundary this page will not cross

PV = p·v is the limit that governs a dry or boundary-lubricated bush, and p is on the PROJECTED area. p = W/(d·L), where d×L is the shadow the bush casts and not the curved surface it has. Using the wrapped area πdL understates the pressure by a factor of π, and it understates the PV by the same — always in the flattering direction. The projected area is not a convenient convention: the load is carried by the arc the shaft is pressed into, with the pressure falling to zero at the sides, and the integral of that distribution over the loaded arc is exactly the load divided by the projected area. Every published PV limit is quoted on it.

There are three limits and the smallest governs. The pressure limit p_max, which needs L ≥ W/(d·p_max). The PV limit, which needs L ≥ W·v/(d·PV_max). And the velocity limit v_max, which is a pass or fail and no amount of length will fix. This page computes all three and reports the longest bush any of them demands, because that is the dimension you can act on. For the worked example — 1,200 N on a 20 mm shaft at 300 rev/min in sintered bronze — the pressure limit asks for 4.4 mm and the PV limit asks for 10.8, so the PV limit governs by a factor of two and a half and a bush sized on pressure alone would have been less than half long enough.

Watch the units: 1 psi·ft/min = 3.5025 × 10⁻⁵ N/mm²·m/s. The PV limits in every US source are in psi·ft/min and the metric ones are in N/mm²·m/s (which is the same as MPa·m/s), and the two differ by a factor of nearly thirty thousand. A PV limit of 50,000 for sintered bronze is 1.75 in metric units. This page prints both for every material and both for your duty. The conversion here is built from the exact definitions of the pound-force, the inch and the foot rather than from a rounded factor. A second unit trap in the same tables: for the sintered metals, the pressure column that matters for a rotating duty is the DYNAMIC one, which is four times lower than the static limit for bronze — 2,000 psi against 8,000 — and the static figure is the one usually quoted.

PV limits are quoted at a stated duty and continuous running derates them. QBC’s sintered bronze technical section is the one source this batch could find that quantifies it: a stated maximum PV of 50,000, and then “for long-time running with no additional lubrication, 20000 should be a limit in selecting loads for various speeds”. That is a factor of 2.5. Applying the same factor to the other ten materials in the table is this site’s extrapolation from one published statement, which is why the duty is a selector rather than an assumption and why the page says whose number it is. The manufacturer’s own continuous rating for your bush takes precedence over anything here.

Length-to-diameter: two published bands, and they disagree. Armoloy gives 0.5 to 1.0 for a plain bush, with the reason that shorter “aids misalignment tolerance”. MITcalc gives 0.3 to 1.0 for a hydrodynamic journal bearing, noting that the load capacity peaks near 0.4. This batch looked for a published band reaching 1.5 and did not find one. What the two agree on is what happens at the top: above about L/d = 1 the thing that governs stops being the mean pressure and becomes the DISTRIBUTION of it, because the shaft slopes and a sloping shaft in a long bush loads one edge. That makes the shaft’s own slope at the bearing the number that decides a long bush. Two short bushes spaced apart is usually a better arrangement than one long one.

Running clearance, and what closes it. A plain bush needs clearance to work at all — somewhere for the lubricant to be and room for the shaft to find its own position — and thermal expansion takes it away. That matters most on a plastic bush, whose coefficient of expansion is five to ten times steel’s: a plastic bush in a steel housing running hot grows INWARDS against a shaft that has hardly moved, and the clearance assembled cold disappears. Nylon does the same thing with moisture, which it absorbs and swells to. Neither appears anywhere in a PV calculation, both are in the bush maker’s fitting instructions, and both are why a plastic bush is specified with more running clearance than a metal one. For the general fit arithmetic see the ISO 286 fit calculator and for the thermal part the thermal effect on fit calculator.

REFUSED: the hydrodynamic case. Everything above is for a dry or boundary-lubricated bush, where the surfaces touch and PV is a proxy for the heat that makes. A hydrodynamic bearing is a completely different calculation: a pressurised oil film separates the surfaces, the load capacity comes from the Sommerfeld number So = F·ψ²/(D·B·η·ω), the answer is a minimum film thickness, and that thickness has to be compared against a limit built from both surfaces’ roughness, the oil filter’s fineness, the shaft’s slope and its waviness. ISO 7902 parts 1 to 3 specify it and Raimondi and Boyd’s charts are the classical solution. None of that is derivable from a PV limit and none of it is guessable, so this page names the boundary and stops. A PV number for a hydrodynamic bearing is not conservative — it is meaningless.

Frequently asked questions

Is the pressure on the projected area or the wrapped area?

Projected — d×L, the shadow the bush casts. Using the wrapped area πdL understates the pressure by a factor of π and the PV by the same, always in the flattering direction. It is not merely a convention: the load is carried by the arc the shaft presses into, with the pressure falling to zero at the sides, and integrating that distribution over the loaded arc gives exactly the load divided by the projected area. Every published PV limit is quoted on the projected area, so using anything else compares two different quantities.

What units are PV limits in?

US sources publish them in psi·ft/min and metric ones in N/mm²·m/s, which is the same as MPa·m/s, and one psi·ft/min is exactly 3.5025 × 10⁻⁵ N/mm²·m/s. So a PV limit of 50,000 for sintered bronze is 1.75 in metric units. This page prints both for every material and for your own duty. Mixing the two is the commonest PV mistake and it is off by a factor of nearly thirty thousand, which at least means it is obvious once you look.

Why does the calculation give three different limits?

Because there are three separate failure modes. The pressure limit is about the material yielding — a sintered bronze’s pores closing, a plastic creeping. The velocity limit is about surface heating at high speed regardless of load, and no amount of bush length fixes it. The PV limit is about the total heat generated per unit of bearing area exceeding what can be conducted away. The first two are independent hurdles; the third is the product. This page computes the minimum bush length each of the first and third demands and reports the longer, then checks the second as a pass or fail.

Do PV limits need derating for continuous running?

Yes, and this is the honest answer about how much: only one of the sources consulted quantifies it. QBC’s sintered bronze technical section gives a maximum PV of 50,000 and then says that “for long-time running with no additional lubrication, 20000 should be a limit in selecting loads for various speeds” — a factor of 2.5, so 0.4. This page offers that as a duty selector and applies the same 0.4 to the other materials, which is an extrapolation from one published statement and is labelled as such. Use the bush maker’s own continuous rating where you have it.

What length-to-diameter ratio should a bush have?

Between about 0.5 and 1.0 on one published band, and 0.3 to 1.0 with the load capacity peaking near 0.4 on another — and this batch could not find a published band reaching 1.5. What matters more than the exact numbers is what changes at the top of the range: above about L/d = 1 the governing consideration stops being the mean pressure and becomes the distribution of it, because the shaft slopes and a long bush then carries its load on one edge. Two short bushes spaced apart is usually better than one long one.

Can I use PV for a hydrodynamic bearing?

No, and this page refuses to pretend otherwise. A hydrodynamic bearing has a pressurised oil film separating the surfaces, so there is no rubbing contact for PV to be a proxy for. Its calculation runs through the Sommerfeld number So = F·ψ²/(D·B·η·ω), the eccentricity ratio and a minimum film thickness compared against a limit built from both surfaces’ roughness, the filter fineness, the shaft’s slope and its waviness — ISO 7902 parts 1 to 3, with Raimondi and Boyd’s charts as the classical solution. A PV number for such a bearing is not conservative, it is meaningless.

What closes a plain bush’s running clearance?

Heat, mostly, and moisture on a nylon bush. A plastic’s coefficient of thermal expansion is five to ten times steel’s, so a plastic bush in a steel housing running hot grows inwards against a shaft that has barely moved and the cold clearance disappears — which is why plastic bushes are specified with a larger running clearance than metal ones and why the maker’s figure should be used rather than a general fit table. Nylon also absorbs water and swells, so a bush that was correct dry can be tight after a week in a humid place. Neither effect appears anywhere in a PV calculation.

Related calculators

References

SDP/SI, General characteristics of plastic bearings and non-metallic bearings and An introduction to porous metallic bearings. The source for the PV, maximum pressure, maximum velocity and maximum temperature columns for the non-metallic materials, and for the DYNAMIC pressure column for the sintered metals, which is the one a rotating duty needs and which is four times lower than the static column for bronze.
QBC Bearings, technical section 2.0, Sintered-Metal Bearings. The second, independent printing of the sintered PV and velocity columns, which agreed with SDP/SI’s exactly, and the source of the only published DERATE this batch could find: “for long-time running with no additional lubrication, 20000 should be a limit in selecting loads for various speeds” against a stated maximum of 50,000 — a factor of 2.5.
Armoloy, Plain bushings (sleeve/journal) — selection, failures and fixes, and MITcalc’s Hydrodynamic plain journal bearings help. The two sources for the length-to-diameter band, and they do not agree: Armoloy gives 0.5 to 1.0 for a dry or boundary-lubricated bush, MITcalc 0.3 to 1.0 with the load capacity peaking near 0.4 for a hydrodynamic one. Neither supports a band reaching 1.5. Both are printed.
ISO 7902 (parts 1 to 3), Hydrodynamic plain journal bearings under steady-state conditions. Cited by number and explicitly OUT OF SCOPE for the plain bearing page, which computes a PV limit and not a film. It is named so the boundary is clear: ISO 7902 needs the Sommerfeld number So = Fψ_eff²/(D·B·η·ω), a relative clearance chosen from its own series, the eccentricity ratio, and a minimum film thickness compared against a limit built from both surfaces’ roughness, the filter fineness, the shaft’s slope and its waviness. None of that is a PV calculation and none of it is guessable.
ISO 15243:2017, Rolling bearings — Damage and failures — Terms, characteristics and causes. Cited by number. It is the document that gives the failure modes this batch keeps naming — surface-initiated and subsurface fatigue, smearing, fretting corrosion, false brinelling — their agreed names, which matters because a defect frequency identifies a LOCATION and not a cause.