Arm length is the capacity
Every other decision in cantilever racking is downstream of this one, and it is the opposite of what buyers expect.
In pallet racking, a longer beam holds less. In cantilever racking the same physics applies with a vengeance, because an arm is supported at one end only. The load hangs out in front of the column, and the further out it hangs, the greater the bending moment at the connection. Published product data states the mechanism plainly: capacity decreases as arm length increases due to greater moment load on the connection.
Here is what that looks like across a single arm family, where nothing changes except the length.
| Arm length | Capacity per arm |
|---|---|
| 12 in | 1,000 lb |
| 18 in | 750 lb |
| 24 in | 600 lb |
| 30 in | 500 lb |
| 36 in | 400 lb |
| 42 in | 350 lb |
| 48 in | 300 lb |
The instinct is to order generous arms so nothing overhangs. Resist it. Material stored across cantilever arms is supported by several arms at once, so the arm only has to reach far enough to support the load stably, not far enough to underrun its full width. Specify from the material's depth and how many uprights it will span, not from a wish for margin. Every inch of unnecessary arm costs capacity and money simultaneously.
Published guidance on the heavier series states the relationship directly: taller heights generally pair with more, shorter arms at a lower per-arm capacity, while shorter heights use fewer, longer arms at a higher per-arm capacity. You are choosing where to spend a fixed structural budget: outward in reach, or upward in levels.
Three capacities, not one
Per arm, per upright and per system. A quote that gives you one of them has told you a third of the answer.
The duty tiers, and what they buy you
Compare like for like. At 36 inch arms, published per-arm capacity in the light series is 400 lb. In the heavy series it is 1,200 lb. Same reach, three times the load, because the column, the base and the connection are all heavier.
| Model | Series | Height x base | Arms | Per arm |
|---|---|---|---|---|
| SM6 | 1000 | 6 ft x 33 in | 6 x 12 in | 1,000 lb |
| SM8 | 1000 | 8 ft x 33 in | 8 x 24 in | 600 lb |
| SMD10 double | 1000 | 10 ft x 78 in | 24 x 36 in | 400 lb |
| SU8 | 4000 | 8 ft x 37 in | 8 x 24 in | 2,000 lb |
| SU10 | 4000 | 10 ft x 49 in | 12 x 36 in | 1,200 lb |
| SU12 | 4000 | 12 ft x 61 in | 16 x 48 in | 1,000 lb |
| SDU10 double | 4000 | 10 ft x 82 in | 24 x 36 in | 1,200 lb |
Look at SM8 and SU8 in that table. Both are 8 feet tall with eight 24 inch arms. One holds 600 lb per arm, the other 2,000 lb. The heavier series costs roughly double for more than triple the capacity, which makes it better value per pound if you need that capacity, and a waste if you do not. Ask every supplier to quote at your arm length, not at their most flattering one.
The base is set by the arm
Not by the height, which is what almost everyone assumes. This is the detail that decides whether the rack stays upright.
A cantilever upright is a lever. Load on the arm tries to rotate the column forward, and what stops it is the base extending far enough in front to resist that rotation. So the base length has to grow with the arm length, because that is what sets the overturning moment.
Published upright data makes this explicit. In one medium-heavy series, base length determines which arms you are allowed to fit:
| Base length | Compatible arms |
|---|---|
| 38 in | 12, 18 and 24 in |
| 50 in | 30 and 36 in |
| 62 in | 42 and 48 in |
A 62 inch base occupies more than five feet of floor in front of the column, and on a double-sided rack the base runs both ways. Published double-sided bases in these ranges are 78 inches and 82 inches. When you price a long-arm configuration, price the aisle you lose to the base as well as the arm, because a rack you cannot get a forklift past is not a saving.
If a supplier quotes 48 inch arms on a 38 inch base, something is wrong: either the arms are not rated for that base or the configuration is not the one they priced. Cross-check the base against the arm on every quote. It is a ten second check and it catches genuine specification errors.
Straight or inclined arms
Identical capacity, different jobs. Choose on what the material does when nobody is holding it.
Note first what the published pricing shows: at every length, straight and inclined arms carry exactly the same rated capacity. A 12 inch arm is 1,000 lb either way. So this is not a strength decision. It is a containment decision.
Stable material
Straight arms
For material that stays where it is put: lumber, sheet goods, bar stock, structural steel and stable unitized loads. The flat presentation makes loading and unloading straightforward from either side and wastes no vertical space.
Anything that rolls
Inclined arms
These tilt the stored material back toward the upright rather than holding it level, so cylindrical or bundled material cannot roll forward during storage and retrieval. The specified use is pipe, round tubing, electrical conduit bundles, bar stock and mixed cylindrical stock. Gravity does the containment.
This is a detail worth knowing because it explains something you will otherwise think is a defect. Published specification is that straight arms include a minimum pitch of 3/8 inch per foot to compensate for deflection under load. The arm is manufactured tilted slightly upward so that when it is loaded and deflects, it reads level. An unloaded cantilever rack that looks like its arms point gently skyward is behaving exactly as designed.
Retaining lips
Both arm types are offered with and without a lip. Published guidance is that a lip provides additional containment for materials that may shift during handling, while versions without lips allow faster loading and unloading when the incline alone provides sufficient control.
Every lip is something a forklift tine or a bundle has to clear on the way in and out. On a fast-moving line that costs seconds on every pick, all day. Specify lips where material genuinely shifts, and leave them off where the incline already holds the load. Buying lips everywhere as a precaution slows the operation permanently to solve a problem you may not have.
Published detail: arms adjust up and down the column on 3 inch centres using a hairpin keeper, so no tools are needed to change a level. That is genuinely useful, because cantilever storage almost always gets re-levelled once real material arrives and you discover the bundles are a different height than the drawing assumed. Plan to adjust once after the first month.
Single sided or double sided
The question is not how much you want to store. It is where the rack is standing.
That is the whole catch, and it is a floor plan question rather than a rack question. A double-sided run consumes its base depth plus a working aisle on each side. Against a wall it is worse than useless, because half of what you bought is unreachable. Measure the aisles before you price the value.
Series, and the interchangeability trap
The most expensive mistake available in this category, and it is stated plainly on the product page in language most buyers skim past.
Published product data says it directly: "Series 1000 uprights, arms, and brace sets are not interchangeable with any other rack series." The same statement appears on the medium-heavy series. Arms will not fit another series' columns, brace sets will not fit another series' uprights, and a bay of one series cannot be extended with components of another. This is not a compatibility inconvenience; it means the series you choose is a decision you live with for the life of the installation.
What the tiers are for
Because components do not cross series, the cheap decision made today caps what the installation can ever become. If your material mix is trending heavier, or you expect to add height, the cost of stepping up a tier now is far less than the cost of running two incompatible systems later. This is the one place in the category where buying ahead of your needs is defensible.
Starter and add-on units
Cantilever is sold as a chain, not as individual bays, and pricing it as individual bays will overpay by a third.
A cantilever run is a line of uprights with material spanning between them. The first bay needs two uprights. Every bay after that needs only one more, because it shares an upright with the bay before it. That is precisely what the starter and add-on split describes: starter units include uprights, arms and braces for one complete bay, while add-on units expand existing bays without duplicating structural components.
| Unit | Type | Arms included |
|---|---|---|
| SM6 | Single starter | 6 x 12 in |
| AM6 | Single add-on | 3 x 12 in |
| SMD10 | Double starter | 24 x 36 in |
| AMD10 | Double add-on | 12 x 36 in |
The commonest ordering error is counting the bays you want and ordering that many units. Draw the run, count the uprights, and the answer is one starter plus however many add-ons complete the line. If the run has a break in it, that break needs its own starter, because a free-standing end needs two uprights again.
The published rationale for this architecture is that facilities can build out their rack system over time. That works only within a series, as section six covered. If you expect to extend, buy the series that will still suit you when you do, and keep a note of exactly which series you installed. That note is worth more in three years than it seems today.
ANSI MH16.3 and the 2025 revision
Cantilever has its own standard, separate from pallet racking, and it changed recently in ways that affect what you should ask for.
Buyers frequently assume ANSI MH16.1 covers all racking. It does not. ANSI MH16.3, the Specification for the Design, Testing and Utilization of Industrial Steel Cantilever Storage Racks, is the governing document for this category. A supplier who answers "MH16.1 compliant" to a cantilever question has told you something about their engineering.
What the 2025 revision changed
Confirmation that the design is to ANSI MH16.3 and which edition. The engineer of record's inspection list, which the current revision requires. And a load plaque or capacity documentation stating per-arm and per-upright capacity at your configured arm length. None of these should be an unusual request.
These are published summaries of the standard's scope and revisions, given so you can ask informed questions. Cantilever rack design, seismic requirements and inspection obligations must be handled by a qualified engineer against the full text of the current standard and your authority having jurisdiction. This guide is educational and is not a structural engineering opinion.
Anchoring, overturning and seismic
A cantilever rack is trying to fall forward all day. What stops it is the base, the anchors and the slab.
Pallet racking fails mostly by being hit. Cantilever racking has a second and more constant threat, because the geometry itself generates an overturning moment whenever the arms are loaded. That is why the base is the size it is, and why anchoring is not optional.
A double-sided rack loaded evenly on both faces largely balances itself. A single-sided rack has all its load on one side by definition, and relies entirely on base length and anchors to stay upright. If you are installing single-sided against a wall, do not treat the wall as structural unless an engineer says it is. The wall stops the rack going backwards. Nothing about it stops the rack going forwards.
What to settle before installation
- Slab thickness and condition. An anchor is only as good as the concrete around it. Cracked slabs, thin slabs and expansion joints running under a base all reduce holding capacity. Establish this before the design, not on the day the drill comes out.
- Anchor specification from the manufacturer or engineer. Anchor count, diameter and embedment are engineered outputs, not site decisions.
- Seismic requirements as a design input. These are geographic and feed into base size, anchor design and sometimes the whole configuration. The 2025 revision specifically addresses overturning percentages for anchorage design, so a current design should reflect that.
- Bracing. Brace sets tie uprights into a rigid run. They are series-specific, they are structural rather than optional, and a run assembled without its full brace set is not the structure that was engineered.
- Outdoor installations. If the rack is going in a yard, snow load and wind become live considerations, and the 2025 revision addresses how snow combines with other loads.
Cantilever sits outside ANSI MH16.1, but the owner-side habits are the same: annual inspection, immediate isolation of damaged structure, and re-evaluation after any reconfiguration. Our pallet racking guide covers those obligations, load plaques and the R-Mark in detail, and most of the discipline transfers directly.
How the load is placed
Two racks with identical components can be safe or overloaded depending purely on where the material sits.
The 2025 revision of the standard added a section on exactly this, with illustrations showing how off-centre load placement affects arm forces. It was added because it is a real and common failure path, and because it is invisible: nothing about a badly placed load looks different from a well placed one until something bends.
Most yards and shops accumulate short remnants, and short remnants end up on the same rack as full lengths because that is where the material lives. Give offcuts their own defined location, ideally on shorter arms or a lower level, and mark it. It costs nothing and it removes the single most likely way this rack becomes overloaded without anyone deciding to overload it.
Two habits that cost nothing. Heavy material belongs on the lower arms, which lowers the centre of gravity of the whole run and reduces the overturning moment. And material pushed back against the column rather than resting near the arm tip reduces the bending moment at the connection for the same weight.
Choosing by material
What we would specify, and why, for the loads that come up most.
Common and costly mistakes
Each is cheap to avoid at quote stage and expensive to discover afterwards.
Ordering the longest arm that fits
Published capacity falls from 1,000 lb at 12 inches to 300 lb at 48 inches, and the long arm costs roughly double. Buy the shortest arm the material genuinely needs.
Assuming base length follows height
It follows arm length. The same upright height is offered in three base lengths, each matched to a band of arm lengths.
Quoting one capacity figure
Per arm, per upright and per system are three different numbers. Ask for all three at your configured arm length.
Comparing series at different arm lengths
An 8 ft unit with 24 inch arms is 600 lb per arm in one series and 2,000 lb in another. Compare like for like or the cheaper quote always wins on paper.
Mixing components between series
Uprights, arms and brace sets are explicitly not interchangeable between series. The series you buy is a long-term commitment.
Ordering a starter for every bay
Bays after the first share an upright. Four starters instead of one starter plus three add-ons costs about 34% more for the same run.
Buying retaining lips everywhere
Lips slow every load and unload. Specify them where material actually shifts, not as blanket insurance.
Thinking a bent-up unloaded arm is a fault
Straight arms carry a minimum pitch of 3/8 inch per foot so they read level under load. Upward pitch when empty is correct.
Treating the wall as structure on a single-sided rack
The wall does not resist the forward overturning moment. Base length and anchors do.
Storing offcuts with full lengths
Short heavy remnants concentrate on one or two arms a load the design spread across five. Give offcuts a defined location.
Loading toward the arm tip
Same weight, greater moment. Push material back toward the column, and put the heavy material on the lower arms.
Asking for MH16.1 compliance
Cantilever is governed by ANSI MH16.3. Ask for that, the edition, and the engineer of record's inspection list.
Frequently asked questions
What comes up most often on quote calls.
Why does a longer arm hold less?
Because an arm is supported at one end only, so the load creates a bending moment at the connection, and that moment grows with the distance from the column. Published product data states it directly: capacity decreases as arm length increases due to greater moment load on the connection.
The scale surprises people. In one published series the same arm design runs 1,000 lb at 12 inches, 600 lb at 24 inches, 400 lb at 36 inches and 300 lb at 48 inches. Four times the reach costs you 70% of the capacity, and the long arm is roughly double the price. Buy the shortest arm the material actually needs.
How long does the base need to be?
The base is matched to the arm length, not to the height, because arm length is what sets the overturning moment trying to rotate the column forward. Published uprights in one medium-heavy series come in three base lengths that dictate which arms you may fit: a 38 inch base takes 12, 18 and 24 inch arms; a 50 inch base takes 30 and 36 inch arms; a 62 inch base takes 42 and 48 inch arms.
Note that the same three base options apply from 6 feet tall right up to 20 feet. If someone offers longer arms without changing the base, cross-check it, because either the arms are not rated for that base or the configuration is not what was priced.
Straight arms or inclined arms?
Choose on whether the material rolls, not on strength, because published capacities are identical at every length. Straight arms suit lumber, sheet goods, bar stock, structural steel and stable unitized loads. Inclined arms tilt the material back toward the upright so cylindrical or bundled stock cannot roll forward, which is why they are specified for pipe, round tubing, conduit bundles and mixed cylindrical stock.
One detail worth knowing: straight arms include a minimum pitch of 3/8 inch per foot to compensate for deflection under load. An unloaded rack whose arms point slightly upward is behaving exactly as designed, not sagging in reverse.
Do I need retaining lips?
Only where material genuinely shifts during handling. Published guidance is that a lip provides additional containment for materials that may shift, while versions without lips allow faster loading and unloading when the incline alone provides sufficient control.
The cost of a lip is not the price, it is the seconds added to every load and unload, permanently. Specify lips selectively rather than as blanket insurance, and let inclined arms do the containment where they can.
Single sided or double sided?
It is a floor plan question. Single sided suits a rack against a wall, with published bases from 33 to 62 inches projecting forward. Double sided suits open floor, with bases of 78 to 82 inches extending both ways, and it nearly doubles the arm count for well under double the price because the column, base and bracing are shared.
The catch is that double sided needs a working aisle on both faces. Against a wall, half of what you bought is unreachable. Measure the aisles before you price the value.
Can I mix components between series?
No. Published product data is explicit that uprights, arms and brace sets are not interchangeable with any other rack series. Arms will not fit another series' columns and brace sets will not fit another series' uprights.
This makes the series choice a long-term commitment rather than a price decision. If your material mix is trending heavier, or you expect to add height later, stepping up a tier now costs far less than running two incompatible systems. Keep a record of which series you installed; it is worth more in three years than it seems today.
What is the difference between a starter and an add-on unit?
A starter includes uprights, arms and braces for one complete bay. An add-on expands an existing bay without duplicating structural components, because bays after the first share an upright with the bay before them.
Order accordingly or you will overpay significantly. Using published Series 1000 single-sided figures, a four-bay run bought as four starters instead of one starter plus three add-ons costs about 34% more. Draw the run, count uprights rather than bays, and remember that a break in the run needs its own starter.
Which standard applies to cantilever racking?
ANSI MH16.3, the Specification for the Design, Testing and Utilization of Industrial Steel Cantilever Storage Racks. This is a separate document from ANSI MH16.1, which governs pallet racking and explicitly excludes cantilever. A supplier answering "MH16.1 compliant" to a cantilever question has told you something useful about their engineering.
Ask for the standard, the edition, the engineer of record's inspection list, and capacity documentation stating per-arm and per-upright figures at your configured arm length.
What changed in the 2025 revision of MH16.3?
Four things worth knowing. Arms now have explicit deflection limits, where previously they only had to avoid deflecting below horizontal and published commentary notes there was not a strict requirement before. The engineer of record must verify design intent and identify the elements requiring inspection, with the standard stating what inspections are needed before the system operates.
Loading requirements were updated, including how pallet load, seismic load and snow load calculations are combined, which matters because cantilever is frequently installed outdoors. And seismic provisions were revised, including overturning load percentages for anchorage design, with the direct analysis method now permitted for stability calculations.
Does a cantilever rack have to be anchored?
Yes, and it matters more here than in pallet racking, because the geometry generates an overturning moment whenever the arms are loaded. A single-sided rack has all its load on one side by definition and depends entirely on base length and anchors to stay upright.
Do not treat a wall as structural on a single-sided installation unless an engineer confirms it. The wall prevents the rack going backwards; nothing about it prevents the rack going forwards. Settle slab thickness and condition before the design, and take anchor count, diameter and embedment from the manufacturer or engineer rather than deciding on site.
Does it matter where on the arm I put the load?
Considerably, and the 2025 revision added a section with illustrations on exactly this, showing how off-centre load placement affects arm forces. Capacity is quoted for the design load position, so material pushed to the outer end of the arm increases the moment for the same weight.
Three habits cover it. Load material back toward the column rather than at the arm tip. Put heavy material on the lower arms, which lowers the centre of gravity and reduces the overturning moment. And spread long material across as many uprights as it reaches, because per-system capacity assumes that sharing.
What is the most likely way my rack becomes overloaded?
Offcuts. Long material spreads its weight across several uprights, which is what the design assumes. A short heavy remnant sits on one or two arms and delivers the whole load where the design expected a share.
Every shop and yard accumulates remnants, and they end up on the same rack as full lengths because that is where the material lives. Give offcuts a defined location, ideally on shorter arms or a lower level, and mark it. It costs nothing and it removes the commonest way a cantilever rack becomes overloaded without anybody deciding to overload it.
Quick-reference spec table
Published figures, verified against manufacturer data and the published standard. Confirm against current documentation before ordering.
| Arm length | Capacity per arm |
|---|---|
| 12 in | 1,000 lb |
| 18 in | 750 lb |
| 24 in | 600 lb |
| 30 in | 500 lb |
| 36 in | 400 lb |
| 42 in | 350 lb |
| 48 in | 300 lb |
| Base | Arms permitted |
|---|---|
| 38 in | 12, 18, 24 in |
| 50 in | 30, 36 in |
| 62 in | 42, 48 in |
| Model | Series | Config | Height x base | Arms | Per arm |
|---|---|---|---|---|---|
| SM6 | 1000 | Single starter | 6 ft x 33 in | 6 x 12 in | 1,000 lb |
| AM6 | 1000 | Single add-on | 6 ft x 33 in | 3 x 12 in | 1,000 lb |
| SM8 | 1000 | Single starter | 8 ft x 33 in | 8 x 24 in | 600 lb |
| SMD10 | 1000 | Double starter | 10 ft x 78 in | 24 x 36 in | 400 lb |
| AMD10 | 1000 | Double add-on | 10 ft x 78 in | 12 x 36 in | 400 lb |
| SU8 | 4000 | Single starter | 8 ft x 37 in | 8 x 24 in | 2,000 lb |
| SU10 | 4000 | Single starter | 10 ft x 49 in | 12 x 36 in | 1,200 lb |
| SU12 | 4000 | Single starter | 12 ft x 61 in | 16 x 48 in | 1,000 lb |
| SDU10 | 4000 | Double starter | 10 ft x 82 in | 24 x 36 in | 1,200 lb |
| Series | Duty | Heights | Per arm at 24 in | Per arm at 36 in |
|---|---|---|---|---|
| 1000 | Medium | 6, 8, 10 ft | 600 lb | 400 lb |
| 2000 | Medium-heavy | 6 to 20 ft | Per configuration | Per configuration |
| 4000 | Heavy | 8, 10, 12 ft shown | 2,000 lb | 1,200 lb |
| Item | Detail |
|---|---|
| Governing standard | ANSI MH16.3, cantilever storage racks |
| Not governed by | ANSI MH16.1, which excludes cantilever |
| 2025 revision, arms | Explicit deflection limits introduced |
| 2025 revision, inspection | Engineer of record identifies elements requiring inspection |
| 2025 revision, loads | Pallet, seismic and snow load combination updated |
| 2025 revision, seismic | Overturning percentages revised; direct analysis method permitted |
| Arm adjustment | 3 in centres, tool-free hairpin keeper |
| Straight arm pitch | Minimum 3/8 in per foot, to offset deflection |
| Component compatibility | Not interchangeable between series |
| Family | Role |
|---|---|
| Single-sided uprights with base | Wall-adjacent runs, base projects forward only |
| Double-sided uprights with base | Open floor runs, arms both faces on a shared column |
| Straight arms | Lumber, sheet, bar stock, structural steel, unitized loads |
| Inclined arms, with or without lip | Pipe, tube, conduit and anything that rolls |
| Brace sets | Tie uprights into a rigid run; structural, not optional |
| Starter units | Uprights, arms and braces for one complete bay |
| Add-on units | Extend a run by sharing the previous upright |
Capacities and prices are as published and subject to change; confirm against the current data sheet for your exact series, arm length and configuration. Per-arm, per-upright and per-system capacities are different figures and all three depend on how the load is placed. Standards material here is educational and is not a structural engineering opinion or a compliance assessment; cantilever design, anchoring, seismic requirements and inspection obligations must be handled by a qualified engineer against the current edition of ANSI MH16.3 and your authority having jurisdiction. Tell us the material, its length and weight, whether it rolls, and whether the rack goes against a wall or in the open, and we will specify it.