01What the basic formula leaves out
The basic rating life is a laboratory figure. It is calculated for a raceway at full hardness, a bearing running at normal temperature, and a load that is constant in size and direction. A machine tool breaks all three assumptions at once, so the correction is not an optional refinement: it is the difference between a rail that lasts five years and one that lasts eighteen months.
The corrected form keeps the same structure and adds three multipliers:
Roller L = ( fh × ft × C / (fw × P) )10/3 × 100 km
Anything that shortens life without being one of these three, such as contamination, poor lubrication or a badly aligned rail, is not in the formula at all. Section 05 covers how to handle it.
02fh: raceway hardness
Rolling contact fatigue assumes the raceway is through-hardened to roughly HRC 58 to 62. Below that, the surface yields earlier and the rating falls quickly, not gently. Induction-hardened shafts, soft machined rails and repaired surfaces are where this factor matters.
| Raceway hardness | fh, roughly | What it means |
|---|---|---|
| HRC 58–62 | 1.0 | The design case. No derating. |
| HRC 55 | about 0.9 | Soft spots start to matter; life falls by roughly 30%. |
| HRC 50 | about 0.7 | A third of the life is gone before the machine has run. |
| HRC 45 or lower | 0.5 or less | Not a bearing surface. Reconsider the material, not the size. |
The percentages above are a working guide for early estimates; the hardness curve in the supplier catalogue is the authority, and it differs by series. Two practical points matter more than the exact number: hardness is measured on the raceway, not on the rail blank, and a rail that was re-machined after hardening may have lost the case depth that gave it the rating.
03ft: running temperature
Above roughly 100 °C the steel itself starts to lose hardness, and the non-metallic parts become the real limit: end caps, wipers, retainers and the grease. In practice a standard block is a 100 °C component regardless of what the raceway could survive.
- Up to 100 °C. ft = 1.0 for a standard block.
- 100 to 150 °C. Derate according to the catalogue curve and confirm that the end caps, seals and grease are rated for it; a standard block is not.
- Above 150 °C. Special heat-resistant specification, metal end caps and a high-temperature lubricant. Expect a different lead time and a different price, and say the temperature on the RFQ.
Temperature also changes the fit. A rail that grows 0.5 mm over a metre while the frame stays cool will either bow or load the mounting bolts; that is a design problem, not a life factor, and it belongs in the mounting-surface discussion.
04fw: load factor for shock and speed
This is the factor that separates a catalogue calculation from a machine. It covers vibration, reversal shocks and the dynamic component that a steady-state load figure never shows. Most catalogues tie it to the speed band, because faster axes are usually also the ones that reverse harder:
| Application | Speed | fw |
|---|---|---|
| Smooth, no shock | V ≤ 15 m/min | 1.0–1.2 |
| Light shock | 15 < V ≤ 60 | 1.2–1.5 |
| Normal machining load | 60 < V ≤ 120 | 1.5–2.0 |
| Heavy shock, press, high speed | V > 120 | 2.0–3.5 |
The band is a starting point, not a rule. A slow press with a hard bottom dead centre deserves 2.0 even though its speed sits in the first row. What raises fw is acceleration and impact, not metres per minute on their own. If the axis reverses direction with a heavy carriage, the reversal is the shock.
05A worked correction
Take the conveyor duty from the mean load page: a 25 size ball rail with C = 24.8 kN and a mean load of 9.64 kN per block, running at 80 m/min in a normal machining environment. The uncorrected figure is:
Corrected fh 1.0, ft 1.0, fw 1.5
L = (24.8 / (1.5 × 9.64))3 × 50 = 252 km
In hours T = 252 × 1000 / (80 × 60) = 52 h of travel
The same hardware, the same load, and the life drops by a factor of 3.4 once the machine's own shocks are included. Fifty-two hours of travel sounds alarming on a continuously running conveyor, and it is the reason conveyors are usually sized one or two steps above the static calculation, with a larger block count rather than a larger rail.
Now suppose the rail was supplied induction hardened to HRC 55 instead of the specified through-hardened raceway, fh = 0.95:
That 14% loss of life per 5 HRC points is why the hardness specification belongs on the drawing, next to the grade and preload, and not in a footnote.
06What sits outside the formula
Three effects are larger than any of these factors on a badly built machine, and none of them appear in the correction:
- Contamination. A single hard particle passing through the raceway can take an order of magnitude off the life. Wipers and covers are covered in dust protection, and no correction factor substitutes for them.
- Lubrication. Running dry does not reduce life linearly; it changes the failure mode to wear. The interval is in relubrication.
- Mounting accuracy. A rail forced into a curved bed loads every block continuously. The corrected life calculation assumes the rail is straight; if it is not, the number is fiction. See mounting surface.
The honest way to use the formula is to calculate it, then ask which of these three is the weakest link in the actual machine, and fix that first.
07Putting it on the RFQ
Load 9.6 kN mean per block, 12 kN peak
Speed 80 m/min, reversal every 4 m
Environment coolant mist, 35 °C ambient
Life target 20,000 h, two shifts
Ask C and L10 with fw stated, and the hardness of the raceway
Asking for the life with the factor stated makes the quotation comparable. Two suppliers can quote the same rail and the same C, and still disagree by a factor of three on life, purely because one applied fw and the other did not.
08Questions we get
Does the catalogue life already include these factors?
No. L10 assumes full hardness, normal temperature and a steady load. The user applies the factors, which is why identical rails can show lives that differ by three times.
Which factor costs the most life in practice?
Usually fw. It covers the machine's own shock and vibration, and a press or a fast reversing axis commonly lands at 1.5 to 2.0, which shortens life by three to eight times on its own.
Can I recover the life by choosing a bigger rail?
Partly. Life follows the cube of the C ratio, so stepping from 25 to 30 with about 1.56 times the C gives roughly 3.8 times the life. It does not fix a hot, dirty or misaligned application.
09Where to go next
The size calculator gives you C from the load; this page turns that into a life figure the machine will actually deliver. The mean load page prepares the P that goes into it, moment load covers the peak check, and accuracy and preload explains why a ZB block starts with a higher internal load.
Standard ball rails for conveyors: HG series. Roller rails for shock-loaded axes: RG series. If the duty is still unclear, send the cycle and we will calculate both the corrected life and the size it implies.
Part of the MS Linear product selection library — the engineering notes behind our linear guideways, ball screws, modules, shafts and bearings. For the full range, browse the MS Linear catalogue or the selection library index.
This page covers the essentials. For the complete selection PDF, full part-number tables and a matched recommendation, contact MS Linear or email [email protected] — we will send the relevant selection PDF for your load case.
Buying Precision Motion Components from China
is Your Best Choice!
Factory-direct quality and worldwide shipping — standard parts leave the factory in 3–10 days, custom strokes and leads in 15–20 days (2–3 weeks). Send your drawing or the original part number and get a quotation within 24 hours.
Cookie preferences
Choose which categories this site may use. Necessary cookies cannot be switched off.
