Obround Slot Milling: Does the Width Stay Constant?

Obround slot milling is the step that decides whether a linkage arm adjusts or fights. Job 260414004 arrived in September 2023 from a maker of linkage, lever and mount components whose previous shop had delivered arms that were, by their own inspection report, correct: the obround slot was the right width, and they could prove it, because they had measured the slot at both ends. The customer's complaint was the opposite of a wrong dimension. The arm slid freely at the beginning and the end of its adjustment and jammed in the middle of it, which is where the adjustment was actually used.

The arms in the photograph are the production run that followed: a forked linkage arm in 1Cr17Ni2 martensitic stainless, roughly 80 mm long, quenched and tempered to the hardness the drawing calls for. One end is a rounded head carrying a large through-bore with a counterbore step and a small radial tapped hole; the middle is an elongated body with an obround adjusting slot and a tapped hole; the other end is a fork of two prongs with a stepped shoulder. Every edge is chamfered, the body is bead blasted to a uniform matte grey, and the part number is laser marked. The bore and its seating face read brighter than the body because they are bored and faced.

What this arm has to do?

Three jobs, and the slot sits between two of them.

The head end pivots or carries a bearing. The large through-bore is the arm's rotation point, the counterbore step gives the bearing or the fastener head a seat, and the small radial tapped hole locks whatever sits in the bore. Because the head is the pivot, the bore is the arm's natural reference: every other feature is positioned from it.

The body adjusts. The obround slot is the travel: a fastener passes through it, the arm slides, and the useful adjustment is wherever the slot is straight. The slot is also the thinnest section of the body, because a slot of any real length leaves a wall on each side of it.

The fork end pulls. The two prongs take a pin and the stepped shoulder sets how deep the mating part sits, so the fork's gap and the shoulder depth are the features that carry the load.

An obround slot connects the pivot to the pull. It is the one feature the customer moves a few dozen times in service, which is why a slot that measures correct and binds is worse than a slot that measures wrong and is caught at goods-in.

Why a slot is not a hole?

A hole has a diameter; a slot has a width and a length

A bore is described by one number, and a gauge either enters it or does not. An obround slot is a shape with two straight sides joined by two semicircular ends, and the width is nominally the same everywhere along it. That is the description on the drawing, and it is also the assumption everyone makes afterwards.

The assumption is what fails. A slot is a width at a position, and the useful question is not what the width is but whether it holds. A slot that is 0.03 mm tight at one station along its length passes a gauge at both ends and holds the fastener tight over the third of its travel the customer actually uses.

The cutter bows, and the slot follows the cutter

Milling an obround slot means taking a small cutter down a long path. The cutter is the least rigid thing in the setup: it is thin enough to fit the slot, and it is hanging out far enough to reach through the part. Feed it along the slot and it bows sideways under the cutting load, so the slot it leaves is not a constant width. The middle of the slot comes out narrower than the ends, which is exactly the pattern the customer described, because the deflection build is smallest where the cutter is entering and leaving the cut and largest where it is cutting a full-length wall.

The second effect runs the same way. A slot narrow enough to matter is cut with a cutter close to the slot width, and a cutter that fills the slot has almost no room to clear chips. Recutting chips load the cutter further, which adds to the deflection and to the variation.

The walls beside the slot are thin, and they move

Cutting the slot removes the material that was holding the body straight, so what remains is a wall on each side of the slot with far less stiffness than the solid body had. The walls can be pushed away by the cutter and spring back behind it, which leaves the slot oversize at that station, or the walls can move in the hardening cycle afterwards, which changes the slot after the last cut has been taken.

That is the second half of the problem and the harder half. A slot cut in the soft state is a slot that still has to survive the heat cycle, and a thin pair of walls on a part whose other end is a thick head will not move by the same amount.

Where you measure a slot decides whether it passes?

The narrowest station governs the travel

The fastener does not average the slot. It has to pass the tightest point anywhere along the length, so the arm's adjustment travel is set by the minimum width, not by the width at the ends and not by the nominal figure on the drawing.

This is why a report that lists the slot as correct and a customer who says the arm jams can both be honest. The report documents a dimension taken at a station, and the customer is describing a length. Both readings are real; only one of them is the requirement.

A slot gauged at its two ends has not been gauged

The ends of an obround slot are the easiest place to measure and the worst place to judge. The ends are where the cutter is entering and leaving, which is where its deflection is least, so they read close to nominal on a slot that is pinched in the middle.

The practical answer is a check that follows the length. A pin or plug gauge sized below the slot's low limit is run through the slot in stages, or the slot is measured at several stations with a gauge that reaches down its length, so the narrowest station is found rather than assumed. On a part that adjusts, that check is the one that matters; the width at the ends is process control, not acceptance.

How the arms were made, step by step?

Step 1 - Rough the arm and leave stock on the reference faces

The arm is roughed from 1Cr17Ni2 bar with the outline brought close to size and a deliberate allowance left on the head faces and the body's outer faces. Those are the faces the arm is referenced from later, and they are cut after the heat cycle rather than before it.

Step 2 - Face the head end and bore the head in the soft state

The head faces are prepared and the large through-bore is drilled and bored while the material is free cutting, which is when the cut is quick and kind to tooling. The head is the arm's reference, so the bore is where the slot will be measured from, not the other way round.

Step 3 - Rough the obround slot and leave the walls stock

The slot is roughed with a cutter sized to leave material on both walls and at the ends rather than brought to width in one pass. The roughing pass does the bulk of the removal, and the finishing pass is what decides the width, which is why the two are separated.

Step 4 - Quench and temper to the drawing's hardness

The rough arm goes through the full quench and temper cycle to give it the hardness the drawing calls for. The thick head and the thin fork respond at different rates, and the slot walls are free to move once the slot exists, so the arm is not at its final shape until the cycle has finished.

Step 5 - Face the head and the body from one setup

The head faces and the body's outer faces are faced after the cycle, in one setup, so the arm's reference surfaces come from a single cut and the slot has something stable to be measured against. This is the step that stops the arm being finished from references the assembly never uses.

Step 6 - Finish the obround slot to width in a stepped pass

The slot is finished with a cutter kept sharp and as short as the part allows, a light radial engagement, and the cut taken in stages along the length rather than in one continuous sweep. Taking the finishing cut in stages is what breaks up the deflection build, because the cutter leaves the cut and recovers between stages instead of loading continuously through the middle of the slot.

Step 7 - Cut the slot's ends and confirm the width follows the length

The semicircular ends are finished to match the slot width, and the width is then checked at several stations along the slot, including the middle, with a gauge that reaches down the length. Checking the ends alone is what produces a passing slot and a jammed arm.

Step 8 - Tap, blast, mark and check the arm as it will be used

The tapped holes are cut and cleaned, the body is bead blasted to remove the surface left by hardening and to give a uniform matte finish, and the part number is laser marked last so the mark is not disturbed by blasting. Release checks cover the bore, the counterbore seat, the fork gap and shoulder, and the slot's travel: the arm is fitted with the fastener it will actually use and slid from end to end before the batch ships.

What the 1,800-piece run measured?

Eighteen hundred arms shipped in 20 days, and the customer reported that the adjustment ran free through its whole travel, the bearing seated flat, and the fork lined up with the mating part without rework.

The fault we found in our own trial run

Our trial arms passed the slot check and jammed on the customer's bench, and the check was the problem as much as the machining was.

Every trial arm had been measured at the two ends of the slot, where the width came out inside tolerance, on a slot that had been finished in one continuous pass. When we gauged the trial arms along the length instead, the middle of the slot was tight, and on some arms tight enough that the service fastener would not pass. The continuous finishing pass had let the cutter deflect through the middle of the cut, and measuring the ends had hidden it.

Three changes went into production. The finishing pass on the slot was split into stages along the length so the cutter recovers between them and the deflection never builds to its full value. The slot's release check was changed from a measurement at the ends to a gauge run along the length that finds the narrowest station. And the arm is now fitted with the fastener it will actually use and slid from one end of the travel to the other before the batch is released, because the customer's complaint was about a movement rather than about a dimension.

Between the trial and production the mid-travel jam disappeared, and the difference between what we measured and what the customer felt stopped existing.

Three ways to produce an obround slot, compared

The middle row is the most common way an obround slot is produced, and it is the one that produces an inspection report and a jammed arm at the same time. The choice is worth making explicitly, because the cost of the first row is a slower finishing pass and the cost of the second is a batch that has to be sorted at the customer's bench.

FAQ

Why does an obround slot measure to width and still bind?

Because the width at the ends of the slot and the narrowest width anywhere along it are two different measurements. A slot milled in one continuous pass is widest where the cutter is entering and leaving and narrowest in the middle, where the cutter has deflected furthest under load. A gauge at either end passes and the fastener still binds partway through the travel, because the fastener has to pass the tightest station, not the average one.

How should an obround slot be measured?

Along its length rather than at its ends. Run a pin or plug gauge sized below the slot's low limit through the slot in stages, or measure the width at several stations with a gauge that reaches down the length, so that the narrowest station is found rather than assumed. Keep the end measurements as process control and make the along-the-length check the one that decides whether the part is accepted.

Does the heat cycle change the slot after it is cut?

It can, and on an arm with a thick head at one end and thin fork prongs at the other it usually does. Cutting the slot removes the material that held the body straight and leaves a wall on each side, so those walls are free to move, while the head is heavy enough to hold its shape. Because the two ends of the same part cool at different rates, the slot is finished from the head as the reference and then confirmed after the cycle, rather than being assumed to have survived it.

Can an obround slot be milled in one pass if the cutter is rigid enough?

A cutter that fits the slot is limited by how much of it can be held in the toolholder, so rigidity runs out before the requirement does. A stiffer cutter does reduce the deflection, and it does not remove the chip-clearing problem that comes with a cutter that fills the slot. Splitting the finishing cut into stages along the length is more reliable than trying to buy the rigidity back, and it does not rely on the cutter being perfect for the whole length of the cut.

What to put on your drawing?

Six lines keep an adjusting slot from passing inspection and failing in service:

1. Give the slot width with a tolerance and state that it applies along the whole length and at every station, not at a nominated section.
2. Give the slot's position from the arm's pivot bore rather than from the outer profile, since the travel is what the bore and the slot have to agree on.
3. State the slot's straight-side length and the end radius, so the end form is a requirement rather than whatever the cutter leaves.
4. Say whether the slot is finished before or after the heat cycle, and if it is finished before, say that the width applies after the cycle.
5. Give the fastener or pin size the slot has to pass, and say that it must pass through the full travel.
6. State the finish and where the marking goes, so blasting does not land on a seating face and the mark is applied after the surface is final.

An obround slot looks like the simplest feature on a linkage arm and behaves like the most demanding one, because it is a width that has to hold over a length on a part that is thin exactly where it has to be accurate. Finish the slot in stages, reference it to the pivot, and check it along its length rather than at its ends, and the arm will adjust the way the drawing promised.