ISO 9001 certification in progress RoHS material control Vietnam country-of-origin documentation rfq@saolamfg.comEmail WhatsApp +84 86 501 7914

Guide

What surface finish and machining marks should a valve stem have so a dynamic O-ring seal lasts?

Short answer

A valve stem is the shaft that moves through the seal of a faucet cartridge, shut-off valve or quarter-turn valve, and its outside surface is where a dynamic O-ring wears or survives. Seal vendors give a narrow roughness window for that surface, warn against both rough and mirror finishes, and advise against soft metals such as brass as the sliding surface, which is the central conflict for a turned lead-free brass stem.

By Saola Manufacturing engineering team · Published · Updated

On this page

What finish do seal vendors recommend for a dynamic sliding surface?

Apple Rubber calls 10 to 20 microinches the most desirable metal surface roughness for dynamic seals, and ERIKS says the rubbing surfaces should be 8 to 16 RMS without longitudinal or circumferential scratches. The two ranges overlap between 10 and 16 microinches.

Converted at 0.0254 µm per microinch (derived), 10 to 20 microinches is 0.25 to 0.5 µm and 8 to 16 is about 0.20 to 0.41 µm, per the two vendor pages (Apple Rubber, ERIKS). ERIKS writes its numbers as RMS, Apple Rubber gives its range in microinches without naming the parameter, and the ERIKS manual pairs 32 RMS with 0.8 µm Ra, so the two are treated as equal for this purpose. Confirm the parameter with your seal supplier before you copy a number to a drawing.

This guide is about the stem or other sliding surface, not the groove. For the groove, squeeze and fill see O-ring groove design for machined parts; the gland tables in the ERIKS manual give the groove surface X as 32 microinches (0.8 µm Ra), which is a groove value and not a stem value.

Dynamic O-ring sliding-surface guidance from Apple Rubber (accessed 2026) and the ERIKS technical manual (undated PDF, about 2005); vendor microinch values quoted as written, roughness terms follow ASME B46.1-2019
PointApple RubberERIKS
Reciprocating dynamic surface10 to 20 microinches desirable8 to 16 RMS, no longitudinal or circumferential scratches
Too smoothA highly polished surface will not hold lubricantFiner finishes reduce lubrication and can cause stick-slip; below 5 microinches not recommended
Too roughNot stated on the pageRougher finishes cause excessive wear
Surfaces named as bestHoned, burnished or hard chrome plated; shot-peened or electropolished hold lubricantHoned, burnished or hard chrome plated
Soft metalsAluminum or brass should generally not be usedAluminum, brass, bronze, monel and some stainless steels should be avoided
Rotary shaftShaft hard, within 0.0005 inch TIR, 16 RMS; gland 32 RMSRotary guidance not used here

Why is a mirror finish as bad as a rough one?

A rough surface wears the O-ring, and a mirror surface leaves no pockets to hold lubricant, so the seal drags, sticks and slips. The usable window is narrow and sits between the two.

ERIKS says finer finishes reduce lubrication to the O-ring and may cause stick-slipping and irregular wear, and that values below 5 microinches wipe too clean for good moving-seal life. Apple Rubber says a highly polished surface is not desirable because it will not hold lubricant, and names shot-peened or electropolished surfaces as ones that hold lubricant in small pockets.

For a faucet cartridge or shut-off valve stem that runs against a lightly lubricated seal, the lesson is to stop polishing. A drawing that asks for a mirror finish on the stem because it looks good is asking for a seal that sticks; a drawing that leaves the finish at the as-turned value may be asking for rapid wear. Both ends of the window are failures, which is why the callout needs a range and not only a maximum.

Why do seal vendors say to avoid brass as the sliding surface?

Both vendors name brass among the soft metals that should generally be avoided for dynamic sealing, because a soft surface scores and wears. That collides with the common brass stem, so the drawing must decide which finishing or design route resolves it.

Apple Rubber says softer metallic surfaces, such as aluminum or brass, should generally not be used for dynamic applications. ERIKS lists aluminum, brass, bronze, monel and some stainless steels as soft metals to avoid for moving surfaces, and it adds that where two metals meet the O-ring groove should go in the softer one so the harder one is less damaged.

The table sets out the routes that follow from those statements. It is derived from the two vendor pages and is not a vendor recommendation; the seal supplier makes the final call for the duty. The guide on dezincification and DZR brass covers the alloy question that sits behind a brass stem in potable water.

Routes for a brass stem under a dynamic O-ring, derived from Apple Rubber (accessed 2026) and ERIKS (about 2005) guidance and not a vendor statement; roughness terms follow ASME B46.1-2019
RouteWhat it changesWatch-outs
Stainless stem, finished by grinding or polishing to the seal windowHarder sliding surface in a material the vendors do not list as soft, except some stainlessERIKS lists some stainless steels as soft; confirm grade and hardness with the seal supplier
Brass stem with a hard coating such as hard chromeBoth vendors name hard chrome among the best surfacesPlating is a partner-plant step; on a drinking-water part ask the certification body whether the coating is covered by the listing
Brass body, harder sleeve or insert where the seal slidesMoves the sliding surface off the soft metalAdds a part and a joint; needs a design change, not a drawing note
Brass stem kept, duty limitedAccepts wear for a low-cycle or static dutyOnly with written seal supplier agreement and a life test
Rotary quarter-turn stemVendor guidance differs from reciprocating: hard shaft, tight runout, smoother shaft than glandApple Rubber gives 16 RMS shaft and 32 RMS gland; check with the seal supplier

Which machining marks are acceptable on the stem?

ERIKS asks for no longitudinal or circumferential scratches on the rubbing surface, and a plain turned surface carries circumferential feed marks by nature. Reading the two together, a stem that runs under a dynamic seal usually needs a finishing step after turning.

That reading is the author's, not a vendor statement: ERIKS does not discuss turned stems, and the conclusion follows from the way a cutting tool leaves feed marks around the circumference. It does not mean every turned stem fails. It means the drawing should say what the surface condition is after the final operation and should not rely on the as-turned finish unless the seal supplier has approved it with a life test.

Feed marks are not the only marks. Burrs and sharp edges on the stem are a cut risk for a passing O-ring: a cross hole, a flat, a slot or a shoulder that the seal crosses needs an edge break. The guide on the break sharp edges drawing note explains the note, and O-ring failure modes shows what a nicked or abraded seal looks like in the field.

Ask also that the stem is free of embedded abrasive, because a ground or polished surface can hold grit and a gritty surface wears the seal. A plain statement on the drawing, "surface free of embedded abrasive and of burrs", gives incoming inspection something to check.

How do you write and check the stem finish on the drawing?

Call out a roughness range with a minimum and a maximum on the named sealing surface, agreed with the seal supplier, and add the lay and edge requirements. A single maximum value cannot protect against a mirror finish.

The values belong to the seal supplier, not to this guide. A note shaped like this works: "Sealing surface between [datum or dimension]: Ra [minimum] to [maximum] µm, final operation [method agreed with the seal supplier], no axial or circumferential scratches, no burrs or embedded abrasive; edges at [feature] broken per [edge note]." The guides on surface finish symbols on a drawing and Ra versus Rz explain the callout and the parameter.

Measurement should follow the same parameter and cutoff as the callout, and the first articles should be measured on the actual stem. Surface roughness measurement covers the stylus method, and roughness, waviness and lay are the three constituents of surface texture in ASME B46.1, which the callout should respect when it names the lay.

Then confirm with the real seal: a short life test of the stem against the production O-ring and lubricant is the check the finish number cannot replace. Record the result on the first article report.

What should the RFQ say about the finishing route?

Name the final operation that sets the finish, who performs it, and how it is verified at what frequency. Turning cuts the stem, and the finishing step that makes the seal window is a separate line on the quote.

The stem is turned at the lead plant on CNC turning equipment with a stock allowance left for the final step. Grinding and any hard coating are arranged through partner plants Saola manages, and the quote states the production site and country of origin. Ask the supplier to list the steps in order and to say which one sets the finish.

Ask for roughness readings with every first article and a stated sampling plan for later lots, and ask for the seal supplier's name and part number on the order so the finish was agreed against a real seal. At a year of ten to twenty thousand stems, the finishing step usually drives cost and consistency more than the turning, so quote it as its own line.

Frequently asked

What Ra should a valve stem have under a dynamic O-ring?

Apple Rubber publishes 10 to 20 microinches and ERIKS publishes 8 to 16 RMS, which is typically 0.2 to 0.5 µm in metric terms (derived conversion). Agree the exact window with your seal supplier and confirm it with a life test.

Can a brass stem work under a dynamic O-ring?

Both vendors advise against brass as a dynamic sliding surface because it is soft. If the design needs brass, treat it as a risk to be proved by test, or change the surface or the design.

Is a polished stem better for sealing?

No. A highly polished surface will not hold lubricant, and ERIKS says surfaces below 5 microinches are not recommended for dynamic O-ring seals.

Does a turned finish meet the dynamic window?

It can, but turning leaves circumferential marks that ERIKS says to avoid on the rubbing surface. Specify the final finishing step and prove the stem against the production seal.

Is the groove finish the same as the stem finish?

No. The groove and the sliding surface have different values; the groove is covered in the guide on O-ring groove design for machined parts.

Sources

  1. Seal Design Guide: Surface finishes, Apple Rubber Products, accessed October 2026
  2. ERIKS Technical Manual, O-ring gland design information (PDF), ERIKS, undated PDF with 2005 file date
  3. ASME B46.1-2019, Surface Texture (Surface Roughness, Waviness and Lay), ASME, 2019, reaffirmed 2026

Have a drawing for this? Send it for a quote.

STEP or PDF, one part or a whole family. You get a unit price at your volume, the tooling cost, and a short manufacturability note.