A scroll plate is not one machining problem. It is four features on the same casting, each with its own tolerance, its own tool and its own failure mode, and the interactions between them decide whether the part passes.
This page organises the work by feature rather than by tool type, because that is how the drawing is read and how a process is planned. Each feature gets a tool family, a key tolerance and the mistake that shows up most often.
The four features are the wrap groove, the bearing bore, the Oldham slot and the face.
The groove is a milling problem with a varying radial engagement, the bore is a reaming or boring problem with a position relationship to the groove, the slot is a positional problem, and the face is a flatness and finish problem.
Most scrap on scroll plates comes from the interaction between the thin wall and the groove depth, not from any single feature being out of tolerance on its own.
The scroll geometry itself is covered on the aluminium scroll compressor plate machining page. This page covers tool selection per feature and the tolerances that decide it.
The four features of a scroll plate
The wrap groove carries the involute profile. Its width, its depth and the radius at the root decide the compression behaviour, and its tolerance is not a single number: the effective width changes with the unwrapped angle because the tool engages more material near the root than near the tip.
The bearing bore locates the shaft. It is usually an H6 or H7 bore with a position and perpendicularity relationship to the wrap, and that relationship is what makes it a machining problem rather than a reaming problem.
The Oldham slot takes the anti-rotation coupling. Its width is a fit and its position is a function of the same datum chain as the bore, so a slot that is on width but off position will assemble and then wear.
The face is the sealing and locating surface. Flatness and surface finish decide how the plate sits against its counterpart, and a face that is flat but rough will leak before it wears.
Read in this order, the drawing becomes a sequence of decisions rather than a list of tolerances: profile, then location, then fit, then flatness.
Tool selection by feature
The table maps each feature to a tool family, the tolerance that drives the choice, and the mistake that appears most often in production.
| Feature | Tool family | Key requirement | Common mistake |
|---|---|---|---|
| Wrap groove | PCD end mill or ball-nose cutter, small diameter, long reach | Groove width and root radius held along the full wrap | Using a tool whose deflection at full depth changes the effective width at the root |
| Bearing bore | PCD reamer or fine boring tool | Roundness and position relative to the wrap datum | Expecting a reamer to correct position that the pre-hole established |
| Oldham slot | PCD slot mill or end mill | Slot width fit and position to the same datum | Cutting the slot in a second setup without repeating the datum |
| Face | PCD face mill | Flatness and Ra | Finishing the face before the bore, then clamping on it to machine the bore |
Two of the four choices are forced by geometry rather than by tolerance. The groove needs a small diameter with a long reach, which means deflection is a design constraint; the bore needs either a multi-edge tool that follows the pre-hole or a single-point tool that can move it.
The other two are forced by the datum chain. The slot and the bore both reference the wrap, so they belong in the same setup if the machine can reach them, or in two setups that repeat the same datum.
Tool family alone does not decide the outcome. A PCD end mill in a weak holder will produce a groove that is on nominal width at the top and narrow at the root, and the drawing will show a taper that no cutter geometry explains.
The wrap groove: where the tolerance hides
The groove is cut along an involute, so the tool's radial engagement changes continuously. Near the tip, the cutter removes a thin chip; near the root, the same feed removes more material and pushes harder against the wall.
That variation has two consequences. The first is deflection: the tool pushes away from the wall more at the root, so the groove narrows where the load is highest. The second is wear: the same region of the cutter does most of the work, so the edge wears unevenly along its length.
Uneven wear produces a characteristic drift. The first parts are on width, then the groove begins to narrow at the root, and the operator compensates with a tool offset that opens the tip. The result is a groove that measures on width at one station and off width at another.
The practical check is to measure the groove width at three unwrapped positions, not at one. A single measurement at the tip will accept a groove that is out of tolerance at the root, which is exactly where the sealing function lives.
If the drift appears within a batch rather than across tools, the cause is usually thermal growth of the tool and the part, not wear.
A warm-up cycle and a stable coolant temperature remove most of it, which is the same logic as the machine-side checks in machine-side requirements before running a PCD reamer.
Thin wall meets deep groove
A scroll plate has a thin wall between adjacent wraps and a deep groove beside it. That combination produces three effects at once, and they add rather than cancel.
The first is clamping distortion. A plate clamped on a face that is still being machined will spring when released, and the groove that was on width under clamp moves with the wall.
The second is cutting force deflection. The wall is thin enough to bend away from the cutter, so the groove becomes wider at the top of the wall and narrower at the base.
The third is residual stress. A die-cast or forged blank carries stress that is released as material is removed, and the release is asymmetric because the groove removes more from one side of the wall than the other.
The three effects are separated by measurement timing rather than by tooling.
A part measured in the fixture answers whether the tool cut correctly; the same part measured after release answers whether the fixture held it; a part measured after a temperature soak answers whether the design is stable.
The control methods for all three, including clamping options and stock allocation, are set out in thin-wall aluminium housing distortion control.
Measuring the features without arguing
Each feature has a measurement that settles the argument, and the argument usually comes from measuring the wrong one.
| Feature | What to measure | What it settles |
|---|---|---|
| Wrap groove | Width at three unwrapped positions, plus profile on a sample | Whether the drift is along the wrap or across the batch |
| Bearing bore | Diameter, roundness, and position to the wrap datum | Whether a size problem is being mistaken for a location problem |
| Oldham slot | Width and position in the same datum frame as the bore | Whether the slot and bore were cut in one setup |
| Face | Flatness and Ra on the released part | Whether clamping, not cutting, produced the error |
The order matters. Position and flatness are measured before size, because a size measurement on a distorted part is a measurement of the distortion.
Measure after release, and after a soak if the part is large enough to hold heat. A plate that measures correctly in the fixture and out of tolerance on the bench is not a tool problem, and treating it as one will consume several tools before the fixture is questioned.
What to send with an enquiry
Four items decide whether a scroll plate enquiry can be answered with a tool recommendation rather than a request for more information.
The wrap parameters come first: pitch, wall thickness, number of turns, groove width and depth, and the root radius. These set the cutter diameter and the reach, and the reach sets the deflection risk.
The material and the process come second: die-cast or forged, alloy, temper, and whether the part has been stress-relieved. A die-cast blank and a forged blank with the same geometry behave differently at the same parameters.
The tolerances come third: bore size and position, slot width and position, face flatness, groove width along the wrap. If the drawing does not separate size from position, that is worth flagging before quoting.
The volume comes fourth, because it decides whether a custom PCD tool is justified or a standard cutter with a controlled process will do.
The tool families used for the scroll features are on the PCD scroll end mill page, the bore tools on the PCD reamer page, and the application context on automotive powertrain. Send the package through the custom cutting tool RFQ.
The scroll plate is one family inside a wider component map: how the six aluminium part families are routed, and which bores decide each, is set out in the automotive aluminium bore tool map.
What published sources say about this
These pages were used for the tool behaviour, the tolerance framework and the capability logic. Each is cited for what that page states, under its own conditions.
- Reamers, machine reamers (toolmaker technical overview): reamer structure, guide part and flute shapes, the reference for what a multi-edge tool holds and what it follows.
- Reaming and fine boring (toolmaker competence page): the two finishing families, including the adjustable single-edge variant used where the bore has to be moved.
- Tools for fine machining bores (toolmaker interview): the single-bladed reamer and the handling requirements of fine boring tools, background to the bore feature choice.
- Dimensioning and Tolerancing (ASME Y14.5): the position, profile and flatness definitions that decide how the four features are called out and gauged.
- What is Process Capability? (NIST/SEMATECH e-Handbook of Statistical Methods): process spread against specification limits, the method behind judging whether a groove drift is a capability failure.
- What are control charts? (NIST/SEMATECH e-Handbook): the time-ordered plot that separates tool wear from warm-up drift.
- Levene test for equality of variances (NIST/SEMATECH e-Handbook): a test for equal spread between two runs, useful when two setups are compared on the same feature.
Outbound references are listed because the tool choices depend on them, not as endorsements. No customer, part number or cycle time is stated on this page.
Frequently asked questions
Which feature should be machined first?
The face, then the bore, then the groove, in most processes, because the bore and groove reference the face and each other.
If the plate is thin, the order changes: rough all features, release stress, then finish in a sequence that keeps the clamping on a finished surface.
Can one tool cut the groove and the slot?
Sometimes, if the widths and depths allow it and the reach is sufficient for the groove without excessive deflection.
In most cases the groove needs a longer, smaller cutter than the slot, and combining them compromises the groove width at the root.
Why does the groove measure on width at the tip and narrow at the root?
Because the radial engagement is higher near the root, so the tool deflects more there and the same compensation produces a narrower groove.
Measure at three positions along the wrap before changing the offset; a single measurement at the tip hides the error.
Does a reamer hold the bore position relative to the wrap?
No. A reamer follows the pre-hole, so the position comes from the operation that made the pre-hole and from the fixture.
If the position has to be corrected after the pre-hole, the route needs a boring pass before reaming.
What is the first thing to check if the plate moves after release?
Whether the clamping was on a surface that had already been finished, and whether the part was measured after a temperature soak.
If both are correct and the distortion remains, the wall thickness and the stock distribution are the next suspects; the checks are in thin-wall distortion control.
Updated 13 November 2026: first publication. Scope: the four scroll plate features, tool selection per feature, the wrap-groove tolerance and its drift, the thin-wall interaction and the measurement set per feature. No customer, part number or cycle time is stated.