I review roughly 300+ polymer component batches a year. Every quarter, at least two come back because the o-ring groove wasn't right—wrong depth, wrong width, wrong corner radius. It costs the vendor rework, and it costs us time.
If you're specifying a groove for a static seal in a trelleborg silicone or nitrile o-ring, here's the checklist I run on every first article.
When This Checklist Applies
This is for static face seals and dynamic rod seals using standard cross-section o-rings (AS568 or ISO 3601). It does not cover dovetail grooves, custom cross-sections, or radial seals in rotating applications. If you're designing for rotating shafts, you need a dedicated dynamic seal calculator.
Five steps. Follow them in order.
Step 1: Verify the O-Ring Cross-Section
Before you touch the groove dimensions, confirm the o-ring itself. I've seen a spec sheet call out a -214 (3.53 mm cross-section) but the groove was cut for a -216 (2.62 mm). The result: zero compression. The seal was just a spacer.
Check:
- Is the o-ring material correct? Brown silicone vs. standard silicone? If it's brown silicone, be aware it's often a high-temperature compound—hardness may differ from standard 70 durometer.
- What is the actual cross-section? Use a calibrated pin gauge or optical comparator. Don't trust the bag label.
- Nominal cross-section (d2) must match the groove standard you're using. AS568 calls out specific d2 values for each dash number.
Honestly, I'm not sure why vendors skip this step. My best guess is they assume the bag is correct. In Q1 2024, we rejected 4% of first deliveries because the o-ring itself was wrong.
Step 2: Calculate Groove Depth—Not Width—First
Here's where most specification errors happen. People start with groove width. Wrong order.
Groove depth determines compression. Compression is what makes the seal work. Depth is the critical dimension.
For a static face seal with a 70 durometer NBR or silicone o-ring:
- Compression ratio target: 20–25% for static seals. Dynamic seals should use 15–20% compression.
- Groove depth (h): For 25% compression, depth = 0.75 × o-ring cross-section (d2).
- Example: 3.53 mm cross-section → 2.65 mm groove depth.
Common mistake: Using the exact same depth for a brown silicone o-ring (which is harder, often 80 durometer) as you would for a standard 70 durometer silicone. Higher hardness needs slightly less compression—around 15–20%—or the seal will never compress fully.
I didn't fully understand this until a $5,000 batch of brown silicone seals failed a leak test. The groove was cut for 25% compression on a harder material. The sealant could not displace into the groove. Every single one leaked.
Step 3: Set Groove Width Based on Swell
Once depth is correct, compute width. Width must accommodate the o-ring after it compresses—the o-ring bulges sideways.
Rule of thumb:
- For static seals: groove width (b) = 1.3 to 1.5 × d2
- For dynamic seals: groove width = 1.5 to 1.6 × d2
Now check the material. If you're deciding between nitrile vs. epdm for an application that contacts oils, nitrile (NBR) swells less in oil than EPDM does. But EPDM swells significantly in oils and degrades quickly—that's a material choice, not a groove adjustment.
However, if you're using EPDM in a water or brake fluid application (where it excels), you still need to account for moisture-induced swelling. I usually bump the width by 0.1–0.2 mm for EPDM in continuous fluid contact. Not much, but enough to prevent the groove from acting like a scissors on the seal.
Groove width is not a tolerance relaxation zone. I've rejected parts where the groove was 0.3 mm over width because the engineer thought 'more room is better.' It's not. Too wide and the o-ring doesn't compress properly—it just rolls around in the groove. Reject.
Step 4: Check the Corner Radius & Surface Finish
Sharp corners in a groove will cut the o-ring during installation or under pressure. But an excessively large radius reduces the effective groove volume.
What I check:
- Static seal groove: max corner radius of 0.2 mm (0.008 in).
- Dynamic seal groove: max radius of 0.1 mm.
- No burrs. None. If I feel a burr with a fingernail, the part fails.
Surface finish: For NBR and silicone static seals, I spec 1.6 µm Ra or better. Rough surfaces cause micro-leakage paths. Brown silicone, being harder, is more sensitive—I'd spec 0.8 µm Ra for the groove bottom.
"The vendor claimed a 1.6 µm Ra finish. When we measured it with a profilometer, we got 2.8 µm. They'd 'visually inspected' it. That $4,000 batch went back."
Step 5: Do the 'Feeler Gauge & Pin' Physical Test
By this step, the drawings are right. But drawings and reality don't always match. I do a physical check on every first article:
- O-ring insertion test: Place the o-ring in the groove. It should sit uniformly, not rock or lift at any point. Use a feeler gauge to check gap uniformity around the circumference.
- Compression check: Use a depth micrometer to measure the installed o-ring height against the mating surface. Difference equals compression. Must be within ±2% of target.
- Gland fill check: The o-ring should not exceed 90% of the groove volume when compressed. If it does, the groove is too small or the o-ring too large. (This is where a too-small groove width + correct depth = a squished seal that fails.)
This step—actually installing the o-ring—catches problems that CAD models never show. I've found that the two grooves on a cover plate were cut from different setups, resulting in a 0.15 mm inconsistency between left and right. The drawing said they were identical. The feeler gauge said no.
Watch Out For These 3 Common Errors
1. Assuming 'Standard' Dimensions are Always Standard
AS568 is a standard. But your vendor may have slightly different tooling calibrations. Always cross-check the first article against your spec, not against a generic table.
2. Forgetting the Tolerance Stack
If the groove depth is at +0.1 mm and the o-ring cross-section is at -0.1 mm, your compression ratio just dropped by 5–6%. In a dynamic seal, that's the difference between sealing and leaking at pressure. Calculate the worst-case compression, not the nominal.
3. Mixing Up Material Swell Behavior
Nitrile vs. epdm is not a 'pick one and adjust the groove later' decision. If you pick EPDM for an oil application, the seal will fail from chemical attack, not groove geometry. The groove won't save you. Conversely, if you pick nitrile for hot water, hydrolysis kills the seal. The material comes first, then the groove dimensions.
To be fair, I get why people rush—production timelines are brutal. But a groove that's off by 0.1 mm can cost $15,000 in rework and three weeks of delay. I've seen it happen. Twice this year.
So: run the checklist. Five steps, every time. Your vendor will hate me for making you check their work. Your production line will thank me.