What Nobody Tells You About SFU1605 Load Capacity

The catalog number is not a payload limit

An SFU1605 typically has a dynamic load rating somewhere around 7 to 8 kN, depending on the nut configuration and manufacturer. That's roughly 700 to 800 kg of force.

Every few weeks, someone reads that number and asks me if they can put 500 kg on an SFU1605.

No. That number is not a payload rating. It's a fatigue life parameter. It tells you how the screw will behave under repeated loading over millions of revolutions. It assumes ideal lubrication, perfect alignment, purely axial load, and no buckling or vibration. Those conditions don't exist in a real machine.

The dynamic rating is one input into a life calculation. It is not the answer to "how much can I put on this screw."

If you remember nothing else from this article, remember that.

What actually limits an SFU1605

In practice, four things limit an SFU1605 long before the catalog rating does.

Buckling. When the screw is pushing the table, it's in compression. A long screw under compression behaves like a column and can buckle sideways. The longer the unsupported length, the lower the buckling load. A 300mm SFU1605 can take a lot of compression. A 1200mm SFU1605 will not.

Critical speed. When the screw spins, it whips above a certain RPM. Longer screws whip at lower speeds. On a long SFU1605 axis, this is usually the limiting factor, not load.

Motor torque. The 5mm lead gives good mechanical advantage, but the motor still has to produce enough torque. Steppers lose torque at speed. If you need to run at 2000 RPM to hit your rapid speed, the motor may not deliver the torque you assumed from the datasheet.

Nut and support rigidity. A standard SFU1605 nut with light preload is fine for general positioning. For a machining axis with changing cutting forces, you may need more preload and a stiffer fixed-end support. That changes the friction and the motor requirement.

Whichever of those four hits first becomes the real limit. It's almost never the dynamic load rating.

Horizontal axis: what the screw actually sees

Let me take a common case. A 40 kg gantry on a horizontal X-axis. Linear guides carry the weight. The screw provides the force to move and accelerate the mass, plus whatever cutting force the process adds.

For acceleration, the basic relationship is:

F = ma

At 1 m/s², that's 40 N. At 5 m/s², that's 200 N. Add friction, cutting force, and some margin, and you might be looking at 400 to 600 N of axial force on the screw.

That is nowhere near the 7 kN dynamic rating. The screw is fine from a pure load standpoint.

The problem is usually something else. It's whether the motor can produce enough torque at the required speed. Or whether the screw is long enough that whipping becomes the limit. Or whether the axis can accelerate fast enough to make the cycle time useful.

I've seen 40 kg gantries run for years on SFU1605. I've also seen much lighter machines struggle because the screw was too long and the motor couldn't spin it fast enough.

The load is rarely the problem. The rest of the system usually is.

Vertical axis: same mass, different calculation

Put the same 40 kg on a vertical Z-axis and everything changes.

Now gravity is acting along the screw axis, constantly. The static load is:

40 × 9.81 = 392 N

If you also accelerate upward at 1 m/s²:

40 × (9.81 + 1) = 432 N

And that's before friction, cutting force, or any safety margin.

A 400 N continuous load on an SFU1605 is still well within the screw's capability. But now the motor has to hold that load when the axis is stopped, and it has to lift it against gravity during upward moves. If the motor is undersized, the axis will drift, lose steps, or fault out.

There's also a safety issue. A vertical axis should have a brake or a counterbalance, not just rely on the ball screw to hold the load. A ball screw is not a safety device. I've seen a spindle drop because someone assumed the screw would hold it. It did—until the coupling slipped.

So when someone asks me if an SFU1605 can handle a vertical load, my first question isn't about the screw. It's about the motor, the brake, and the counterbalance.

The 5mm lead is a trade-off, not a feature

The 5mm lead gives the motor a lot of mechanical advantage. That's why SFU1605 works well on small machines with modest motors. The thrust per unit torque is high:

F ≈ 2πTη / L

With a 5mm lead and a reasonable efficiency, you get roughly 1130 N of thrust for every 1 N·m of motor torque. That's excellent for a small screw.

But the trade-off is speed. To move at 10 m/min, the screw has to spin at 2000 RPM. To move at 15 m/min, it needs 3000 RPM. On a stepper motor, torque at 2000 RPM is often half or less of the holding torque. On a longer screw, the critical speed might be below 3000 RPM anyway.

So the 5mm lead is great for thrust and bad for speed. Whether that's the right trade-off depends on the machine. A small router cutting wood at 5 m/min with moderate acceleration will be perfectly happy. A machine that needs to rapid at 20 m/min on a 1200mm axis will not.

That's why I don't recommend the lead until I know the speed requirement and the screw length.

Screw length is where the design usually breaks

I can't overstate this. A short SFU1605 and a long SFU1605 are almost different products.

On a 400mm axis with fixed-floating supports, an SFU1605 is a very capable screw. It will run at 2000 to 3000 RPM without complaint, carry a few hundred newtons comfortably, and last a long time.

On a 1200mm axis, the same screw has a critical speed somewhere around 2000 to 2500 RPM. That might be fine if you only need 8 to 10 m/min. If you want 15 m/min, you're asking for 3000 RPM and the screw will whip.

It also sags more. It deflects more under load. It's more sensitive to alignment. The buckling load drops. The whole mechanical picture changes.

So before answering "can SFU1605 handle this load," I want to know the unsupported length between supports. Not the overall length. The unsupported length. That's what determines critical speed and buckling.

A lot of customers send me the overall length because that's what's in the catalog. The unsupported length is what actually matters.

Preload can eat your motor margin

A standard SFU1605 nut comes with light preload or no preload. If you need more stiffness, you can order a preloaded nut.

But preload increases friction. That means more torque required from the motor, more heat generated in the nut, and potentially shorter lubrication intervals.

On a small screw with a modest motor, adding preload can push the motor over its limit. I've seen builders add preload to fix a stiffness complaint and end up with an axis that overheats and loses position after an hour of running. The preload solved one problem and created two others.

Preload should be selected based on what the axis actually needs. If the machine is a router cutting wood, a standard preload is usually fine. If it's a machining axis with changing cutting forces, higher preload may be justified—but the motor and thermal management have to be sized accordingly.

Don't add preload just because there's backlash. Backlash has several possible causes, and preload is only one fix.

The linear guides carry the load, not the screw

This is one of the most common misunderstandings I see.

A ball screw is a drive element. It pushes and pulls the table. It does not carry the weight of the table. The linear guides do that.

If the machine has a heavy overhung load, the guides and the mounting plate take the moment. The ball screw only sees the axial force. If the guides are undersized or misaligned, the screw will feel it—as side loading, increased friction, and vibration—but the fix is not a bigger screw. It's better guides or better alignment.

I've had customers ask if they should upgrade from SFU1605 to SFU2005 because the table "feels heavy." The real problem was that the guide blocks were too small and the mounting plate was flexing. A bigger screw wouldn't have fixed it.

So when sizing an SFU1605, check the guide system at the same time. The screw is only one part of the axis.

What I'd actually check before saying yes

If you send me an SFU1605 inquiry, here's what I want to know before I say it will work:

• Screw length and unsupported length

• Horizontal or vertical

• Moving mass

• Maximum speed

• Required acceleration

• Cutting force or external load

• Motor type and torque curve

• Nut preload requirement

• Support arrangement

• Duty cycle

• Required service life

With that information, I can check the axial force, the buckling load, the critical speed, and the motor torque. Then I can tell you whether the SFU1605 is adequate or whether you need to step up to SFU2005 or a larger lead.

Without it, I'm just quoting a part number.

When SFU2005 starts to make sense

If the SFU1605 is close to its limits, SFU2005 is the natural next step. It's the same 5mm lead, but 20mm diameter. That gives you:

• Higher buckling resistance

• Higher critical speed

• More stiffness

• More load capacity

But it also brings trade-offs. Larger nut, larger support blocks, more installation space, more rotating inertia, and more cost. The motor may need to be sized differently because the reflected inertia changes.

So I don't recommend SFU2005 just because the load is "a bit heavy." I recommend it when one of the SFU1605 limits is actually being exceeded—usually critical speed on a long axis, or buckling on a compression-loaded axis, or stiffness on a machining axis with significant cutting forces.

For a short, horizontal, light-duty axis, SFU1605 is usually the right choice and SFU2005 is overkill.

The short answer

There isn't one.

SFU1605 can handle a 40 kg horizontal gantry with proper guides and a well-matched motor. It can handle a 100 kg table if the guides carry the weight and the acceleration is modest. It can also fail on a lighter machine if the screw is too long, the motor is undersized, or the alignment is bad.

The number you need is not the payload. It's the axial force the screw actually sees during the worst part of the cycle. Calculate that. Check it against buckling, critical speed, motor torque, and nut life. Then decide.

That's the difference between picking a screw and designing an axis.

If you want a real answer for your machine, send the application details. A few minutes of calculation is much cheaper than finding out after the machine is built.