Is Your NC Servo Feeder Sized for Short Feeds?

The Window the Press Actually Leaves You

A press cycle is not available for feeding. The slide occupies part of it, the pilots need the strip held, and the feed has to finish before the pilots enter.

At 200 strokes per minute the full cycle is 0.30 seconds. On a typical progressive die the feed window is closer to 0.12 seconds.

That window is the only number that matters for sizing. A feeder quoted at a maximum speed will still miss if the window is short, because the strip has to accelerate from rest and stop again inside it.

Three things shorten the window: higher stroke rate, longer die engagement, and pilot release timing that gives the strip less free time than the die drawing suggests.

Three Numbers That Size the Feed

You need feed length, stroke rate and available window. Everything else in a sizing conversation is derived from those three.

Average feed speed is simply the feed length divided by the window. It says nothing about whether the move is possible.

Peak speed is what the rolls actually have to reach. With a triangular profile, where the feeder accelerates and decelerates with no constant-speed segment, peak speed is twice the average.

Required acceleration is what decides grip. It comes from peak speed divided by the acceleration time, which is half the window in a triangular profile.

That third number is the one buyers never ask for and the one that determines whether the strip slips.

A feeder with generous torque but modest acceleration will lose the feed, because torque is not what holds the strip.

Worked Example: 60 mm at 200 SPM

Take a 60 mm feed on a 200 SPM press with a 0.12 second window. Average speed is 0.50 m/s and the triangular peak is 1.00 m/s.

Acceleration time is half the window, 0.06 seconds, so the required acceleration is 16.7 m/s², or about 1.7 g.

That is achievable, but notice what happened: the profile has no constant-speed segment at all. The strip is accelerating or decelerating for the entire move.

Any grip loss therefore lands directly in the feed length. There is no plateau where the strip is simply travelling, so there is no part of the move that forgives a slip.

Sensitivity: What Changes When the Feed Gets Shorter

Halving the feed length at a fixed stroke rate does not halve the difficulty. It roughly doubles the required acceleration, because the window stays the same while the distance shrinks.

Doubling the stroke rate at a fixed feed length is worse. The window halves, so average speed doubles and required acceleration goes up by four.

That is why a line that runs comfortably at 120 SPM on long feeds can become marginal when the same press is pushed to 200 SPM on a short pitch. Nothing changed except the two numbers nobody re-checked.

Material matters too. A heavier strip needs more force to accelerate, and force is delivered through the roll surface.

On 3 mm steel the same acceleration that suits 1 mm strip can exceed what the rolls can transmit.

Roll inertia is the term the formula leaves out. A heavier roll set takes longer to reach speed, and that extra time comes out of the window rather than out of the feed length.

Servo tuning matters for the same reason. A loop tuned for smoothness instead of response adds milliseconds at the start of every move, and milliseconds are what a short feed cannot spare.

Tension from the payoff side counts too. If the loop runs tight, the feeder is accelerating the coil as well as the strip, and the load on the roll surface rises without any change to the feed length.

Where the Formula Stops Predicting

The arithmetic above is a first filter, not a guarantee. Four things make it optimistic.

It assumes perfect grip. Real rolls transmit less force as oil film builds and as the surface glazes. The formula has no term for lubricant.

It assumes an ideal profile. Servo tuning, roll inertia and belt or coupling compliance all add time that the calculation does not see.

It assumes a clean strip. Scale, burrs and pre-pierced holes reduce contact area, and contact area is what transmits acceleration.

It assumes a fixed window. Window depends on die design, and a die revision can take 20 ms off it without anyone updating the sizing sheet.

One honest scope limit remains. Below roughly a 15 mm feed at high stroke rates, the feed stop being the constraint and the die becomes the constraint.

At that point the correct move is usually a different process, not a bigger feeder.

What to Specify When the Arithmetic Says No

If the required acceleration comes out above what the rolls can hold, there are four levers and they are not equally cheap.

• Lengthen the window by changing pilot release timing or die engagement, which costs nothing but a die revision.
• Increase roll contact length so the same acceleration is spread over more gripping surface.
• Use a roll surface that holds grip with less pressure, rather than simply raising pressure.
• Split the feed into two shorter moves if the die allows, which trades stroke rate for acceleration.
• Reduce the stroke rate, which is the option nobody wants and the one that always works.

FANTY has been building feeding equipment for twelve years with a 370-person team, and the sizing sheet is the first thing we ask a customer to fill in, because most oversizing starts with a missing window figure.

The rated feed length range of each frame is listed on our coil line product range.

What to Measure on the First Run

Verify the sizing with measurements rather than with the calculation sheet.

Measure actual window. Trigger on the crank and time how long the strip is free. It is usually shorter than the die drawing claims.

Measure fifty consecutive feeds. Hold the spread to ±0.05 mm, which is the accuracy these units are specified at, and check that the error is random rather than drifting.

Watch the first move after a stop. Cold rolls and settled oil give a different result from the hundredth stroke, and that difference is your real grip margin.

Repeat at the top rate. Run at the fastest rate the die will allow, not the rate you plan to run. Margins disappear at the top of the range.

The wider question of where feed error comes from, across the whole line rather than the feeder alone, is set out in the coil line guide library.

Loop control on the same line is the other half of the timing problem, and it is covered in how much loop depth a feeder needs.