Precision Sinker EDM
When Wire EDM Can't Reach It and Milling Can't Hold It — Sinker EDM Is the Answer
Blind pockets, complex internal cavities, and sharp internal corners in hardened steel and carbide require a process that applies zero cutting force — electrode erosion, not mechanical cutting, defines the geometry.
Carbide · Hardened Tool Steel · Complex Internal Features · Prototype Through Production
Zero
Cutting Force
±0.0002"
Cavity Tolerance
Graphite / Cu
Electrode Materials
Ra 8–32
µin Surface Finish
What Sinker EDM Is
How Sinker EDM Works and Why It's the Right Process for Internal Geometry That Other Methods Can't Produce
Sinker EDM (also called ram EDM or die sinking) erodes workpiece material by discharging controlled electrical sparks between a shaped electrode and the workpiece across a small gap — typically 0.001"–0.003" — filled with dielectric oil. The electrode never contacts the workpiece. Material removal happens through localized melting and vaporization at each spark discharge, and the geometry produced in the workpiece is the inverse of the electrode shape. No mechanical cutting force is applied to the workpiece at any point in the process.
The zero-cutting-force characteristic is what makes sinker EDM the correct process for features that other methods cannot produce without distortion: blind pockets in hardened steel and carbide, complex internal cavities with sharp corners, and features that would deflect or fracture under the cutting forces of milling or turning. The electrode geometry — machined from graphite or copper to exact form — defines the cavity, and the spark discharge removes material to match it. The workpiece hardness is irrelevant to the EDM process; carbide at HV 1,500 and hardened tool steel at HRC 65 machine at comparable material removal rates.
A sinker EDM process for a precision cavity typically uses two to three electrodes in sequence: a roughing electrode that removes the bulk of material at high material removal rate with a coarser surface finish, followed by one or two finishing electrodes machined to closer tolerance that bring the cavity to final dimension and surface condition. The rough-to-finish electrode progression is what allows the process to hold ±0.0002" cavity tolerance without consuming finishing electrode life on rough material removal. Each electrode transition requires re-referencing the datum to maintain positional accuracy across the electrode sequence.
The Engineering Behind It
Why Sinker EDM Process Variables Determine Whether a Cavity Meets Tolerance — or Misses It
Sinker EDM is a thermal erosion process, not a mechanical cutting process. Material removal happens through localized spark discharges at the gap between electrode and workpiece — each discharge melts and vaporizes a small volume of material, and the accumulated effect of thousands of discharges per second produces the cavity geometry. The precision of the result depends on electrode dimensional accuracy, gap control, dielectric condition, flushing adequacy, and electrode wear compensation.
The recast layer — a thin zone of resolidified material at the cavity surface — is an EDM-specific concern that does not exist in machining processes. Recast layer depth ranges from 0.0003"–0.002" depending on the electrical parameters used: high material removal rate settings produce deeper recast layers, finishing settings produce minimal recast but at lower removal rates. For fatigue-critical applications, the recast layer must be removed — its altered microstructure and residual tensile stresses reduce fatigue life. Achieving tight tolerance sinker EDM requires managing:
- Electrode accuracy: The electrode must be machined to exact form before EDM begins — dimensional error in the electrode transfers directly to the cavity. Graphite electrodes are machined on high-speed CNC milling centers; copper electrodes are typically ground or turned. The electrode form tolerance must be tighter than the cavity tolerance to leave margin for the spark gap and any dimensional compensation required.
- Electrode wear and compensation: Electrodes erode during the EDM process. Graphite wears at roughly 10–30% of workpiece material removal volume; copper wears at 5–15%. Electrode wear changes the cavity depth if not accounted for — the electrode must be overcut to the correct depth accounting for predicted wear, or wear must be compensated by in-process depth monitoring. Multiple finishing electrodes may be required to achieve final depth within tolerance.
- Gap and orbital motion: The spark gap (0.001"–0.003" for finishing passes) controls cavity size relative to electrode size. Orbital motion — programmed XY movement of the electrode during sinking — opens the gap for flushing in deep cavities and allows the electrode to produce a cavity slightly larger than its physical dimensions. Orbital motion amplitude must be precisely controlled: it directly affects the finished cavity size.
- Dielectric flushing: Contaminated dielectric with suspended eroded particles re-arcs — the particles create secondary discharge paths that produce erratic spark distribution and uneven cavity surfaces. In deep cavities (depth:width ratio greater than 2:1), dielectric must be pressure-flushed through the electrode or through the workpiece fixture to maintain adequate particle removal. Flushing inadequacy is the most common cause of poor surface finish in sinker EDM.
- Datum control across electrodes: When multiple electrodes are used in sequence, each electrode change requires re-establishing the datum reference to maintain positional accuracy. Electrode change without datum re-verification is the most common cause of positional error between roughing and finishing passes.
What Actually Matters
The Variables That Define Sinker EDM Precision
Sinker EDM failures are process failures — wrong electrode material, incorrect parameters, inadequate flushing, or missed datum verification. These six variables are where cavity tolerance is made or lost.
Electrode Material Selection
Graphite and copper have fundamentally different wear rates and machinability — the wrong choice adds errors before EDM starts
Graphite is the standard electrode material for most sinker EDM applications: it machines faster and more cleanly than copper, handles higher current densities, and produces good surface finish at typical finishing parameters. Copper has a lower wear rate (5–15% versus graphite's 10–30% of workpiece erosion volume) and is preferred for applications where electrode wear directly affects cavity depth accuracy — such as very deep narrow features where wear cannot be predicted reliably. Copper electrodes require grinding or turning rather than milling. Mixing electrode materials between roughing and finishing without adjusting parameters produces inconsistent results because the gap characteristics change between materials.
Electrode Accuracy and Form Tolerance
The cavity is only as accurate as the electrode — form error in the electrode transfers directly to the part
The sinker EDM process cannot improve on electrode accuracy. Form error in the electrode — a radius that's 0.001" too small, a wall that's not perpendicular to the base — produces the identical error in the cavity. For a cavity held to ±0.0002", the electrode must be machined to tighter tolerance than that to leave margin for spark gap variation and any orbital motion effect. Electrode qualification (measuring the electrode form before EDM begins) is not optional for precision work — it's the process control point that prevents machining a cavity that's already out of tolerance before the EDM starts.
Spark Parameters and Recast Layer
Higher material removal rate means deeper recast — for fatigue applications, recast layer depth is a design constraint
Sinker EDM electrical parameters (pulse on-time, peak current, gap voltage) control both the material removal rate and the recast layer depth. Roughing settings that remove material at maximum rate produce recast layers of 0.001"–0.002"; finishing settings at lower energy produce 0.0003"–0.0005" recast. The recast layer is re-solidified material with altered microstructure and residual tensile stress — it is not the same as the base material. For tooling and components in fatigue-critical applications, recast must be removed after EDM: by abrasive finishing, by light grinding, or by chemical etching. Leaving recast in place on a fatigue-loaded surface reduces service life predictably.
Dielectric Flushing Strategy
Inadequate flushing in deep cavities is the most common cause of poor surface finish and erratic spark distribution
Dielectric oil carries eroded workpiece and electrode particles out of the spark gap. When particle concentration in the gap rises above a threshold, re-arcing occurs — the particles create secondary discharge paths that produce uneven material removal and a rough, pitted surface finish that cannot be improved without re-machining the cavity. In cavities with a depth:width ratio greater than 2:1, passive dielectric flow is insufficient and pressure flushing is required: either through-electrode flushing (dielectric pumped through holes in the electrode), through-fixture flushing (dielectric injected through the workpiece fixture), or jet flushing aimed at the gap entry. Flushing strategy must be designed before electrode fabrication to incorporate any flushing holes into the electrode form.
Orbital Motion and Cavity Sizing
Orbital motion amplitude directly sets the final cavity size — it must match the target oversize exactly
Orbital motion — programmed XY movement of the electrode during sinking — serves two purposes: it opens the gap for flushing, and it allows the electrode to produce a cavity larger than its own dimensions by the orbital radius. A 0.005" orbital radius produces a cavity 0.010" larger in the XY plane than the electrode. For a cavity held to ±0.0002", the orbital amplitude must be set with corresponding precision — an error in orbital setting transfers directly to cavity oversize. The finishing orbital radius must be verified against the target cavity dimension before final passes, not assumed from the program. Orbital motion does not affect cavity depth in the Z axis.
Datum Reference and Electrode Change Protocol
Positional accuracy between roughing and finishing passes depends on re-establishing the datum after every electrode change
When a sinker EDM process uses multiple electrodes in sequence — roughing electrode, then one or two finishing electrodes — each electrode change requires re-establishing the XYZ datum reference relative to the workpiece before cutting begins. The electrode change sequence (remove roughing electrode, clean spindle, install finishing electrode, re-reference datum, verify position) must be completed in full for every electrode swap. Skipping the datum verification step on the assumption that the machine retained position is the most common cause of positional mismatch between rough and finishing passes — the cavity is in the right shape but in the wrong location on the part.
Failure Diagnosis
When Sinker EDM Produces a Wrong Part, the Cause Is Specific and Correctable
Oversized cavities, poor surface finish, positional errors, and in-service cracking from sinker EDM are not machine failures — they're process variable failures. Each symptom points to a specific cause.
-
Internal corners rounded instead of sharpCause: Electrode corner wear during roughing passes reducing the corner form, or orbital motion radius too large for the required corner geometry. Internal corners in sinker EDM cannot be sharper than the electrode corner after wear — the corner erodes first because it has the highest current density. For very sharp internal corners (radius ≤ 0.005"), plan for a dedicated finishing electrode with the correct corner geometry, apply minimal orbital motion on the corner pass, and account for electrode corner wear in the electrode form tolerance. Using the same electrode for roughing and corner finishing guarantees rounded corners — two-electrode sequence is required for tight internal corner radii.
-
Cavity depth consistently short of targetCause: Electrode wear not compensated in the depth program, or Z-axis datum lost between electrode changes. Electrode wear removes material from the electrode tip as EDM progresses — if the program drives the electrode to a fixed Z depth without compensating for electrode shortening, the cavity will be shallower than intended by the wear amount. For graphite at 20% wear ratio on a ±0.0002" depth tolerance, this requires compensation on any cavity deeper than 0.001". Verify the electrode length before and after roughing to quantify wear, adjust the finishing electrode Z target accordingly, and re-verify the workpiece datum Z reference after each electrode change.
-
Rough or pitted cavity surface finishCause: Inadequate dielectric flushing allowing particle concentration to rise above the threshold for stable arc discharge. The symptom — a rough, randomly pitted surface with no regular pattern — is distinctive: it's not feed marks, it's re-arc damage from secondary discharges in contaminated dielectric. In deep cavities, the solution is pressure flushing (through-electrode or through-fixture), not simply increasing flow rate of the ambient bath. If the cavity geometry does not allow flushing holes in the electrode, consider reverse flushing (suction through the gap) or jet flushing aimed at the gap entry. Also check dielectric filter condition — a saturated filter recirculates rather than removes particles.
-
Cavity position displaced from datumCause: Datum reference not re-established after electrode change, or workpiece movement between electrode changes from fixture inadequacy. When the finishing electrode is installed and the datum is assumed rather than verified, any error in the electrode change setup — slightly different Z height in the toolholder, minor XY offset — displaces the finishing cavity from the roughing cavity position. On a cavity with positional tolerance of ±0.0005", an electrode change without datum re-verification can consume the entire positional budget in a single setup error. Verify XYZ position of every electrode at the datum reference point before beginning each EDM sequence.
-
Workpiece cracking after EDM at or near the cavityCause: Residual tensile stress in the recast layer combining with existing part residual stress from prior heat treatment, particularly in highly alloyed tool steels (D2, M2, H13) where residual stress after hardening is already elevated. The recast layer from sinker EDM carries residual tensile stress — in hardened steel, this can initiate cracking at stress concentrations immediately post-EDM or under light service load. The risk is higher in thin sections, at sharp corners, and in materials with limited ductility at the as-hardened condition. For fatigue-critical components, recast layer removal after EDM is required — confirm via acid etch (nital etch for steel reveals recast as a white, unetched layer) or metallographic cross-section to verify complete recast removal before releasing parts.
-
Batch hardness or surface variation across production cavitiesCause: Dielectric condition degrading across a production batch, or electrode wear changing parameters without corresponding program adjustment. As a dielectric bath accumulates eroded particles across a production run, the effective gap characteristics change — more frequent re-arcing occurs, material removal rate drops, and surface finish degrades. In production sinker EDM, dielectric condition must be monitored via conductivity measurement (target range 5–20 µS/cm for oil dielectric applications) and filtered or replaced at defined intervals. Also check electrode wear data at each part — if wear is accelerating across the batch, a consumable or dielectric problem is developing, not a machine problem.
How CPI Applies This
Precision Sinker EDM at CPI
Sinker EDM is the right tool for blind pockets and complex internal cavities in hardened components — features that milling can't produce after heat treat without risking distortion, and that wire EDM can't reach because they're enclosed. Because sinker EDM applies zero cutting force and material hardness doesn't affect removal rate, it works equally well on a freshly hardened D2 die block at HRC 62 as on annealed steel. At CPI, sinker EDM is typically the last operation on a component that has already been machined, heat treated, and (where needed) ground to reference dimensions. Having all those steps in-house means the process sequence is designed before the first cut — not improvised after heat treat.
We machine blind pockets, complex internal cavities, and sharp-corner features in hardened tool steels, carbide, and engineering alloys. Our work includes tooling cavities for production applications, prototype features for development programs, and replacement components for customers whose previous supplier produced cavities with corner rounding, positional error, or surface finish that didn't meet specification. When a customer brings us a complete part that needs sinker EDM as one of several operations, we coordinate the full sequence from raw stock to finished component.
We work with OEM tooling engineers and precision component buyers across the automotive and general industrial sectors.
Internal Feature That Milling or Wire EDM Can't Produce?
Need a Blind Cavity or Complex Internal Feature Machined to a Tolerance That Requires Real EDM Process Control?
Tell us about your component — material, cavity geometry, corner radius requirement, depth tolerance, surface finish spec, and whether the application is fatigue-critical. We'll tell you what the process plan looks like and whether it's within our capability.
Request a Quote →