LONGQIAO Technical Team | CNC Motion Components, Jinan Longqiao Electromechanical Equipment Co., Ltd. | Published August 22, 2026
تختلف القيم التصنيفية للحمل ودرجات الدقة وحسابات العمر الافتراضي باختلاف السلسلة والحجم اللذين يتم اختيارهما في النهاية. يتناول هذا الدليل طريقة الاختيار. يجب التأكد من الحجم النهائي بالرجوع إلى البيانات التصنيفية الخاصة بالسلسلة المحددة التي تطلبها.
Heavy duty describes a load case, not a rail family. Profile rail handles moment loading and holds accuracy; supported round rail handles long spans on imperfect bases at lower cost. Which one suits a loaded axis depends on moment direction, unsupported span, and how flat your mounting surface really is.
Machine builders arrive at this term with one number in mind, usually the mass of the moving assembly. That number is the least informative of the four things that decide the rail.
Static and dynamic load. What the axis carries, plus what acceleration and cutting forces add to it. Most builders estimate this reasonably well.
Moment load. What the load does when it is not sitting directly over the bearings. A gantry head mounted ahead of its carriage applies a pitching moment; a heavy fixture offset to one side applies a rolling moment. This is where axes actually fail, and it is routinely left out of the calculation.
Unsupported span. How far the rail runs between points where the machine structure carries it. A short heavily loaded axis and a long lightly loaded axis are different engineering problems that happen to share a search term.
Mounting surface quality. How flat and parallel the surface you bolt to actually is, in the condition it will be in after welding, painting and assembly. This variable decides whether the rail delivers the performance in its catalogue or a fraction of it.
Get the second and fourth of those wrong and no amount of upsizing rescues the design.
A linear bearing block is not equally strong in every direction. Downward load, upward lift, side load and the three moments each have their own rating, and on a profile rail those ratings differ substantially from one another.
Picture a router gantry with the spindle carried forward of the Z carriage. At rest, the blocks see mostly downward load. During a cut, the cutting force acts at the tool tip, some distance below and ahead of the bearing centres, and that offset multiplies into a moment. The block nearest the tool takes compression, the far block takes lift. Lift capacity is the number that governs, and it is not the number most people looked up.
Two practical consequences:
ISO 12090 covers linear rolling bearings as a product family, and the reason directional ratings exist at all is that the raceway geometry carries load differently depending on where it comes from.
Deflection under load does not grow in proportion to length. In standard beam theory, the deflection of a simply supported beam under a distributed load rises with the fourth power of span. Double the unsupported length and the sag goes up by a factor of sixteen, with no change in load at all.
This is why long axes fail differently from heavy axes. An unsupported round shaft on end blocks is the classic case: it works on a short machine, and the same design at twice the length sags visibly under its own moving carriage.
Three ways the industry deals with it, and they are genuinely different approaches rather than variations on one:
Continuous support along the whole length. Supported round rail bolts a shaft to an aluminium base that runs the full length of the axis, and that base transfers load into the machine structure continuously. The shaft no longer behaves as a beam spanning between end points. This is the entire reason SBR and TBR formats exist.
A machine structure stiff enough to carry the rail. Profile rail bolts to a machined surface and relies on that surface for its rigidity. The rail is not a beam; the machine base is. This works extremely well when the base is a proper casting or a machined weldment, and poorly when it is not.
Intermediate supports. Adding support points along an otherwise spanning rail, which helps, but introduces a new requirement that all the support points be coplanar.
| Profile rail (HGR type) | Supported round rail (SBR / TBR type) | |
| Load path | Recirculating balls in ground raceways on a square profile | Ball bushing or slider on a round shaft, shaft bolted to a continuous support |
| Moment capacity | High in all three moment directions | Limited, particularly in roll |
| الصلابة | Higher for a given envelope | Lower, but continuously supported over long runs |
| Accuracy potential | Ground raceways support the higher precision grade conventions | Suited to general machine building rather than precision grades |
| Long spans | Depends entirely on the base being flat and continuous | Handles long, unsupported machine structures because the support extrusion carries the shaft |
| Mounting surface demand | High. Flatness and parallelism errors transfer straight into the blocks | Forgiving. The support base tolerates a less perfect surface |
| Installation | Reference edge, torque sequence, parallelism checks | Bolt down, shim if needed, less sensitive to error |
| Contamination | Sensitive. Chips and dust reach the raceway if unsealed | Also sensitive, but a simpler geometry to wipe and re-lubricate |
| Retrofit into an existing frame | Usually needs the mounting surface machined | Often bolts to an as-built frame |
| Relative cost | Higher, plus base machining cost | Lower, and lower installation cost |
| Best fit | Cantilevered heads, cutting loads, accuracy-driven axes | Long-stroke, long-span, heavy but well-centred loads |
The choice is not a quality ranking. Profile rail on a frame that was never machined flat performs worse than supported round rail on the same frame, because the profile rail’s rigidity depends on a surface it is not getting.
Profile rail transmits its mounting surface into the block. If the surface waves, the block follows the wave, and a preloaded block forced to follow a wave carries internal load before your payload arrives.
The symptoms are recognisable. Running effort that varies with position along the stroke. Two parallel rails that feel fine individually and bind when the carriage bridges them. Bearing life far below what the calculation promised, with wear concentrated at particular positions rather than spread along the rail.
Where this bites hardest is on fabricated frames. A welded steel frame moves during welding and keeps moving as residual stress relieves. A frame that measured acceptably in the fixture will not necessarily measure acceptably after it has been off the fixture for a week. Builders who machine the rail mounting faces after welding get profile rail performance; builders who bolt to as-welded surfaces do not, regardless of what they paid for the rail.
The misconception worth naming directly: that a heavier rail series compensates for a poor mounting surface. It does the opposite. A larger, stiffer rail conforms less willingly to an uneven base, so more of the mismatch ends up as internal preload in the bearings instead of being absorbed as elastic deflection in the rail.
If machining the base is not in the budget, that is a legitimate constraint and it should drive the rail choice rather than being discovered afterwards. Supported round rail exists for exactly this situation.
Adding blocks is the cheapest capacity increase available, and it is usually the right first move on a moment-limited axis.
Spacing them further apart is cheaper still, and more effective. Moment load per block falls as the bearing spacing grows, so a carriage designed with the blocks pushed out to the corners of the available envelope can often use a smaller series than a carriage with the blocks clustered near the middle.
There is a limit, and it arrives sooner than people expect. Two parallel rails carrying four blocks form a statically over-constrained system. Every error in rail parallelism, height difference and mounting flatness has to be absorbed somewhere, and the somewhere is the bearings. On a well-machined base this is fine and standard practice. On a marginal base, four blocks on two rails can run worse than three blocks arranged to allow one degree of freedom to float.
An edge case for gantry builders specifically: a gantry driven from one side twists under acceleration, and that twist shows up as yaw at the tool. It reads like a rail rigidity problem and it is usually a drive symmetry problem. Upsizing the rail treats the symptom at considerable expense. Adding a second drive, or moving the drive closer to the centre of resistance, treats the cause.
HGR profile linear guide. Ball-type rolling structure with three-side ground raceways, supplied as rail and block. Blocks and rails of the same series interchange, which matters more than it sounds when you are holding spares for a fleet of machines or replacing a damaged block years after the build. Custom models are available, the warranty is 1.5 years, and each order ships with a machinery test report and video outgoing inspection. This is the format for cantilevered heads, cutting loads, and any axis where accuracy is the reason the machine exists.
SBR supported rail. Higher load capacity and rigidity for long-span, long-stroke, heavy-duty axes, economical and straightforward to install. This is the format for the case this article keeps returning to: a long axis on a frame that will not be machined to profile-rail standards.
TBR supported rail. Supported guidance with higher running stability and accuracy than unsupported shafting, sitting between the two above.
Linear shaft rod. Used with sliding bearings, heat treated, with controlled diameter and surface finish. Appropriate where the load is light and the geometry simple, and out of place on anything described as heavy duty.
What we do not publish, and will not estimate: rail widths, dynamic and static load ratings, moment ratings, accuracy grades, preload classes and permissible span figures. Those numbers exist in the series data for whatever you order, and they are the numbers your calculation stands on. Send the axis length, moving mass, loadoffset from the bearing plane, orientation and duty cycle, and we will come back with a series recommendation and a quotation. There is no published price list.
That is a different starting point from most of what ranks for this term. Catalogue pages built around a single proprietary rail format will recommend that format, and marketplace listings supply a size and a price with no path from your load case to either. The comparison above is the part that determines whether you are buying the right category before you argue about the size within it.
Linear guide accuracy is conventionally described by grade designations following JIS B 1192 and the equivalent DIN conventions, running from normal through high, precision, super precision and ultra precision. Two things about those grades are worth knowing before you specify one.
First, grade describes manufacturing accuracy of the rail and block, not the accuracy your machine will achieve. Once the rail is bolted to an imperfect surface, the installed accuracy is dominated by the surface. Specifying a high grade onto a poorly prepared base is money spent on a property you then discard.
Second, preload class and accuracy grade are separate selections that interact. Higher preload raises rigidity and reduces deflection under load, which is what a heavily loaded axis wants. It also raises friction, generates more heat, and reduces calculated life. On a heavy axis the correct preload is a deliberate trade rather than a maximum.
Steps four and five are the ones that get skipped, and they are the ones that decide whether the axis works.
Brinelling from shock. Impact loads on a stationary bearing indent the raceway. Machines that crash, or that are transported without the axes clamped, arrive with damage that no amount of lubrication reverses. Clamp the axes for shipping.
Contamination under load. Heavier loads mean higher contact stress, and a contaminated raceway under high contact stress fails quickly. Sealing and wiper condition matter more on a loaded axis, not less.
Lubrication starvation. Load raises the demand on the lubricant film at exactly the moment the film is hardest to maintain. A heavy axis on a forgotten lubrication schedule is a short-lived axis.
Parallelism drift. Rails that were parallel at commissioning may not be after a year of thermal cycling on a fabricated frame. On heavy machines this is worth measuring at service intervals rather than assuming.
Wrong failure attributed. A binding axis gets diagnosed as an undersized rail more often than as a misaligned one. Before upsizing anything, disconnect the drive and move the carriage by hand along the full stroke. Effort that varies with position is an alignment problem, and a bigger rail will not fix it.
A: Nothing intrinsic to the rail. It is a match between the rail’s load and moment ratings and your actual load case, including offsets. A rail that is heavy duty for a centred vertical load may be undersized for the same load on a cantilevered head.
A: No. Profile rail is more rigid and holds accuracy better, but it depends on a flat, parallel mounting surface. On a long axis bolted to a frame that will not be machined, supported round rail usually delivers better real-world performance.
A: That depends on load, orientation and acceptable deflection, and the answer changes very rapidly with length because deflection scales with the fourth power of span. Continuous support removes the question, which is why long axes generally use it.
A: On a moment-limited axis, widen the block spacing first, add blocks second, and upsize last. Upsizing buys capacity in every direction to solve a problem in one, at the highest cost.
A: For the rail to perform as rated, yes. Fitting profile rail to as-welded surfaces produces variable running effort and shortened life. If machining is not possible, that constraint should point you toward supported round rail.
A: Only as high as your mounting surface can deliver. Grade describes the rail as manufactured; installed accuracy is governed by what you bolt it to.
A: Blocks and rails within the same series interchange. Across series or suppliers, do not assume it, even where dimensions appear to match, because raceway geometry and preload class may not.
A: Almost certainly not. Disconnect the drive and push the carriage by hand. Position-dependent effort indicates rail parallelism or mounting flatness, not capacity.
A: It depends on series, length, block type and quantity, so it is quoted rather than listed. Send your axis length, moving mass and load offsets and we will size it and price it together.
Most sizing requests arrive as a rail length and a weight, which is enough information to guess and not enough to answer. A request that gets a real recommendation contains the axis length and required stroke, the moving mass, the distance from the bearing mounting plane to the load centre in each direction, the orientation of the axis, the process forces and where they act, the duty cycle, and one line about the mounting surface and whether it will be machined.
The last item is the one nobody includes and the one that changes the answer most often.
One further thing worth scheduling rather than discovering: on fabricated frames, re-measure rail parallelism after the machine has been assembled and running for a few weeks. Residual stress in a weldment relieves over time, and a rail set that was correct on commissioning day is a rail set that was correct on commissioning day. Whether it is still correct is a measurement, not an assumption.
واتساب
