Battery cooling plates are one of the newer applications I have been seeing in deep hole drilling. They are basically aluminum plates with a network of drilled channels that coolant flows through to keep EV battery cells at the right temperature. I have also written about EV motor shafts and hot runner manifolds — all part of the same EV trend, but each with different challenges.
The concept is similar to a hot runner manifold — a pattern of intersecting holes in a plate — but the scale and the material make it different.
Cooling Plate Design Types
Battery cooling plates come in several configurations depending on the cell format and thermal requirements. Here is what I see in the shop:
| Design Type | Cell Format | Channel Pattern | Typical Thickness | Hole Diameter Range | Channel Count |
|---|---|---|---|---|---|
| Prismatic cell plate | Rectangular pouch or prismatic cells | Parallel straight channels | 15 - 25 mm | 4 - 10 mm | 6 - 20 per plate |
| Cylindrical cell plate | 18650 or 21700 round cells | Serpentine or spiral patterns | 18 - 30 mm | 6 - 12 mm | 4 - 12 per plate |
| Cold plate with manifold | Any cell type | Header and branch channels | 20 - 35 mm | 6 - 16 mm | 8 - 30 per plate |
| Thin profile plate | Pouch cells | Single-pass straight channels | 10 - 18 mm | 3 - 6 mm | 4 - 10 per plate |
The prismatic cell plate is the most common in my experience. The channels are straight-through with cross-drilled intersections at each end to form a manifold. The cylindrical cell plates need curved or serpentine channels that follow the cell layout, which adds complexity because each hole needs a different drilling angle.
Aluminum Changes the Rules
Most of my deep hole drilling experience is in steel. Aluminum is another world. The material is softer, the chips are stickier, and the thermal expansion is three times that of steel.
For battery cooling plates in 6061-T6 or 6063 aluminum, I run:
| Parameter | Value | Notes |
|---|---|---|
| Cutting speed | 150 - 250 m/min | Higher speeds improve chip breakage |
| Feed rate | 0.06 - 0.12 mm/rev | Increase feed if chips are stringy |
| Coolant pressure | 600 - 900 psi | Lower than steel — flow rate matters more |
| Coolant flow rate | 30 - 60 L/min per tool | Flush chips out, not just cool |
| Coolant type | Emulsion with high lubricity | 8-10% concentration for aluminum |
The high cutting speed is the main difference. Aluminum lets you run fast — 200 m/min is comfortable — but the feed needs to be high enough to break chips. At low feed, aluminum produces long stringy chips that pack up and score the channel walls.
I have found that running a slightly higher feed than I would use in steel at the same diameter gives better chip control. If the chips are coming out as fine needles, I bump the feed until they switch to short broken segments.
Chip Control Is the Big Challenge
Aluminum chips are different from steel chips. They’re softer, more ductile, and they tend to pack together under pressure. In a deep hole drilling operation, packed aluminum chips can seize the drill faster than steel chips will.
I run pecking cycles on aluminum battery cooling plates even at moderate depths. A full retract every 100mm clears the chips and lets coolant wash the hole clean. It adds cycle time but I’ve broken fewer drills since I switched to this approach.
The other trick is coolant pressure. Aluminum doesn’t need the high pressure that steel does — 600 psi is usually enough. But the flow rate matters more. I need enough volume to flush the chips out, not just enough pressure to push coolant through the tool.
Flatness and Distortion
Battery cooling plates are thin — typically 15-25mm thick with a grid of holes drilled through them. After drilling, the plate can distort from the residual stress.
I’ve seen plates that measured flat before drilling and had 0.5mm of bow afterward. The fix is to alternate the drilling sequence so the stress builds up evenly. Instead of drilling all the holes from one side, I drill half from one side and half from the other.
For very thin plates (under 15mm), I clamp the plate against a backup support during drilling. The support prevents the plate from deflecting under the thrust of the drill, which also helps keep the hole straight.
Leak Testing
Every battery cooling plate gets leak tested. The channels need to hold coolant pressure without leaking into the battery cavity. A leak that is too small to see can cause a failure in the field.
I use two leak testing methods depending on the production phase. During process development, I use helium leak testing because it detects leaks down to 10^-9 mbar-L/s. In production, I use pressure decay testing because it is faster and the operators can run it without special equipment.
| Leak Test Method | Sensitivity | Cycle Time | Equipment Cost | Best For |
|---|---|---|---|---|
| Pressure decay (air) | 10^-2 mbar-L/s | 30 - 60 seconds | $2,000 - $5,000 | Production leak check |
| Helium mass spectrometry | 10^-9 mbar-L/s | 60 - 180 seconds | $30,000 - $60,000 | Process validation |
| Dye penetrant | Visible cracks only | 10 - 30 minutes | $100 - $500 | Quick visual check |
| Ultrasonic leak detector | 10^-3 mbar-L/s | 10 - 30 seconds | $500 - $2,000 | Portable field check |
I pressure-test plates at 150% of operating pressure. For a cooling plate rated at 3 bar operating pressure, I test at 4.5 bar. The test fixture seals against the channel openings with O-rings, and I hold pressure for 30 seconds minimum. Any pressure drop means a leak — usually at an intersection point where two drilled holes meet.
The most common leak location is where a cross-drilled hole intersects a main channel. If the intersection is not clean, the O-ring cannot seal. I deburr every intersection with a carbide burr before testing. It adds a step, but it catches the leaks before the plate goes to assembly.
I track leak test results by plate serial number and correlate them with the drilling parameters. If I see a cluster of leaks from a particular tool or shift, I investigate the cause. I have found that worn guide bushings cause burrs at intersections, which leads to leak failures. Replacing the guide bushing at the right interval eliminated the leak cluster.
Quality Requirements
Battery cooling plates have tighter quality requirements than most of the manifold work I have done. Here are the typical specifications:
| Requirement | Typical Tolerance | Why It Matters |
|---|---|---|
| Hole diameter | +/- 0.05 mm | Coolant flow rate uniformity across channels |
| Hole straightness | 0.1 mm per 100 mm | Consistent wall thickness between channels |
| Surface finish inside channel | Ra 1.6 max | Pressure drop and particle generation |
| Intersection burr height | 0.05 mm max | Leak sealing and debris in coolant |
| Plate flatness after drilling | 0.2 mm total | Battery cell contact and thermal interface |
| Burr at channel openings | 0.1 mm max | O-ring seal surface damage |
The flatness requirement is the hardest to meet in my experience. I have covered plate distortion management in the machine vibration troubleshooting article, because vibration during drilling contributes to uneven stress relief in thin plates. The coolant pressure and flow parameters I use are also similar to those discussed in the coolant energy comparison — for aluminum cooling plates, the lower pressure range is usually sufficient.
What’s Different from Traditional Channel Drilling
| Aspect | Mold/Manifold (Steel) | Battery Plate (Aluminum) |
|---|---|---|
| Material | H13, P20, 4140 | 6061, 6063 aluminum |
| Speed | 60-80 m/min | 150-250 m/min |
| Feed | 0.04-0.08 mm/rev | 0.06-0.12 mm/rev |
| Coolant pressure | 1200-1800 psi | 600-900 psi |
| Primary risk | Tool wear | Chip packing |
| Distortion risk | Low | Moderate |
The learning curve is real if you are used to steel. My first aluminum battery plate job was harder than I expected because I was running parameters that were too conservative. Once I sped up and increased the feed, the process stabilized.
Key Takeaways
- Battery cooling plates fall into four design types: prismatic cell plates, cylindrical cell plates, cold plates with manifolds, and thin profile plates — each needs a different drilling approach.
- Aluminum drilling parameters differ significantly from steel: cutting speed of 150-250 m/min with coolant pressure at 600-900 psi, where flow rate matters more than pressure for chip evacuation.
- Chip packing is the primary risk in aluminum plates — I run pecking cycles with full retracts every 100mm to clear chips and prevent drill seizure.
- The most common leak location is at cross-drilled intersections, and deburring every intersection before leak testing catches most failures before assembly.
- I use pressure decay testing in production and helium mass spectrometry for process validation, with pressure tested at 150% of operating pressure.
- Flatness after drilling must be held to 0.2 mm total — alternating the drilling sequence prevents distortion from uneven residual stress.
- Prismatic cell plates with straight-through channels are the most common design, while cylindrical cell plates with serpentine patterns add complexity from angled drilling.