A 0.5 mm error on the shaft bore measurement selects the wrong DIN seal size. I’ve watched technicians measure every dimension perfectly on the bench, then order the wrong seal because the shaft they measured was worn 0.3 mm undersize and nobody caught it. The textbook approach — clean parts, good lighting, calibrated instruments — works fine when you have all three. Most replacement scenarios give you none of them.
The measurements that matter most happen in the field with corroded components, limited access, and a maintenance window closing in hours. That reality demands a different technique than any measurement guide will show you. Here is how to get it right.
The Eight Dimensions That Define a Seal Replacement
Before picking up a caliper, know exactly which dimensions you need. Missing one sends you back to the parts counter.
Every mechanical seal replacement requires eight measurements:
- Shaft bore diameter (d1) — the seal’s primary size designation per DIN EN 12756. This is the single most consequential measurement.
- Seat outside diameter (D) — determines gland plate fit and seal chamber compatibility.
- Seat thickness (L1) — affects total stack length inside the stuffing box.
- Spring free length (Lf) — measured with the seal removed and springs fully relaxed.
- Seal installed working length (L) — the dimension most technicians get wrong (more on this below).
- Stuffing box bore diameter (D2) — internal diameter of the seal chamber.
- Stuffing box depth (L2) — measured from the chamber face to the back wall.
- Shaft runout (TIR) — Total Indicator Reading. Not a seal dimension, but exceeding 0.05–0.10 mm tolerance causes face separation within days regardless of seal quality.
These eight form a constraint equation: the stuffing box depth must equal seat thickness plus working length plus 1–2 mm clearance. Violate this relationship and the gland plate bottoms out or the spring loses compression. Either way, the seal fails at startup.
Before you begin measuring, make sure you have identified your seal type — component, cartridge, or bellows — because the measurement procedure differs for each.
Shaft and Bore Measurements
The shaft bore diameter drives the entire seal selection. Get this wrong and nothing else matters.
Shaft Diameter
Measure the shaft at the seal location, not at an exposed section. Take readings at three points along the shaft sleeve area: front, middle, and back. Rotate the shaft 90 degrees and repeat. Six readings minimum. The goal is to detect taper and ovality, not just diameter.
For a new shaft, standard tolerance is +/- 0.025 mm (+/- 0.001 inch). A worn shaft will read below nominal. Record the smallest reading — that is the shaft’s current functional diameter, and it determines whether the existing seal size still fits.
A between-size reading is the most common source of ordering errors. If your measurement falls between two standard nominal sizes, always round to the nearest standard size and verify against the seal manufacturer’s bore tolerance of +0.0/+0.1 mm. When in doubt, call the supplier with your actual mechanical seal measurement rather than guessing which standard size to pick.
Seal Chamber Bore
Use an inside micrometer or telescoping gauge for the stuffing box bore (D2). Calipers lack the reach and accuracy for this measurement. Take readings at two depths — near the opening and near the back — to detect taper from corrosion or erosion.
The most common mistake I see is measuring the bore with a buildup of scale or corrosion still in place. Clean the chamber first, or your reading will be undersize by the thickness of the deposit. A wire brush and some patience save a return trip.
Seal Chamber Depth and Gland Dimensions
Miss the seal chamber depth and the new seal will not physically fit — even if every other dimension is correct.
Stuffing Box Depth (L2)
Use a depth micrometer from the stuffing box face to the bottom of the chamber. This is straightforward on a clean, disassembled pump. On an assembled pump with the seal removed, check that no gasket material or debris is sitting at the bottom — that will give you a false shallow reading.
Remember the constraint equation: L2 must accommodate seat thickness (L1) plus working length (L) plus 1–2 mm clearance. If the depth is too shallow, the gland plate will bottom out before achieving proper spring compression. If too deep, the spring may not fully load the faces.
Gland Bolt Pattern and Pilot Diameter
While you have the gland off, measure the bolt circle diameter and bolt hole size. Also measure the gland pilot diameter — the register that centers the gland in the stuffing box. A worn or corroded pilot means the gland won’t center the seal properly, and no amount of precise seal measurement will compensate for a cocked gland.
Working Length: The Measurement You Cannot Take in the Field
Working length is the most misunderstood measurement in mechanical seal specification, and for good reason — there is almost no reliable method to determine it from a worn seal in the field.
The working length (L) is the installed compressed height of the seal assembly. Most seals require this dimension within +/- 0.030 inch (0.76 mm). That tolerance leaves zero room for estimation.
Here is the problem: working length is a function of spring compression, which changes as the seal wears. Measuring a worn seal’s installed height tells you where the seal ended up, not where it started. The original working length must come from the factory documentation — the seal drawing, the installation manual, or the manufacturer’s spec sheet.
When I pull a seal that is too worn to reference, I follow this sequence:
- Record the manufacturer name and model number stamped on the seal or gland.
- Check the pump OEM manual for seal specifications.
- Contact the seal manufacturer directly with the pump model and serial number.
- If all else fails, measure the spring free length and calculate backward using the manufacturer’s published compression ratio.
This is one reason cartridge seals have grown so popular — they arrive pre-set to the correct working length, eliminating this measurement entirely. But cartridge designs still make up less than half the installed base, so the measurement skill remains essential.
If you skip this step, you will be back in a week replacing a seal that either ran with too little face load or bottomed out on installation.
Measuring Worn and Corroded Components
Corroded shafts, pitted bores, and seal faces with uneven wear — that is the reality of field measurement. The technique for these conditions changes completely from clean-bench procedure.
Why Worn Parts Give Misleading Readings
Each mechanical seal component typically has two or three critical tight-tolerance dimensions — roundness, concentricity, perpendicularity, and press fits. Worn or corroded parts that have been abraded or broken will not yield dimensions identical to the original OEM parts. Measuring them and assuming the readings match the original specification leads to ordering a seal sized to the worn condition, not the design condition.
Seal face wear compounds this problem. Wear is almost always uneven because thermal and pressure-induced distortions do not exactly cancel out. A worn face measured with a micrometer may read 0.003 inch thinner on one side than the other. Taking a single measurement and assuming uniform wear gives a false average.
Field Compensation Techniques
When the shaft shows visible wear or corrosion:
- Measure at an unworn section of the shaft (outside the seal area) to establish the original nominal diameter. Compare against the worn zone reading.
- If the wear exceeds 0.05 mm from nominal, the shaft likely needs sleeving or replacement. Order the seal to the original shaft size, not the worn size.
- For corroded stuffing box bores, scrape or brush the worst deposits and take a minimum of four readings at 90-degree intervals. The largest reading is closest to the original bore diameter.
Available seal face wear length is typically 1/16 to 3/16 inch. If the remaining face thickness is near the lower limit, the seal cannot be re-measured meaningfully — it needs replacement regardless of what the dimensions read.
When to Send the Seal to the Supplier
Some field measurements are simply not accurate enough to stake a replacement order on. If the seal is severely corroded, the shaft is scored, or you cannot locate the working length specification, package the entire seal assembly and ship it to the supplier. A competent seal manufacturer can reverse-engineer the correct dimensions from the hardware in hand — something no set of field measurements from worn parts can reliably duplicate.
The most common mistake I see is technicians spending an hour measuring a destroyed seal to three decimal places, then ordering based on those numbers. A corroded seal face is not a measurement reference — it is scrap metal with memories.
From Measurements to a Seal Order
Accurate measurements are half the job. Translating them into the correct ordering specification decides whether you make one trip to the parts counter or two.
Recording Format
Record every measurement in a consistent format that maps directly to the seal specification:
- Shaft diameter: actual measurement + nominal standard size
- Seal bore (d1): measurement from shaft or seal ID
- Stuffing box bore (D2): measurement + condition notes
- Stuffing box depth (L2): measurement + clearance check against constraint equation
- Working length (L): source document reference (not a field measurement)
- Seat OD, thickness, spring free length: direct measurements
Note the condition of the shaft surface. Shaft runout should be checked with a dial indicator before installing any new seal — if runout exceeds 0.05 mm, fix the shaft condition first. A perfect seal on a bent shaft fails faster than a mediocre seal on a straight one.
Standard Sizes vs. Custom
Most industrial seals follow standard DIN EN 12756 size increments. If your shaft diameter measurement lands on a standard size and the stuffing box dimensions are within normal range, you can order a standard replacement.
When measurements do not match any standard configuration — oversized bores from corrosion, non-standard shaft sleeves, unusual chamber depths — that is when custom-machined seals become the practical solution. Provide the supplier with all eight measurements, condition notes, and process data (fluid, temperature, pressure). A complete measurement package with condition notes gets you the right seal on the first order; partial data with guesswork gets you a phone call asking for the dimensions you skipped.
Next Steps After Measuring
Get the measurements right and the ordering becomes straightforward. Get them wrong and you wait for a second shipment while the pump sits idle.
Before submitting the order, run the constraint equation one more time: stuffing box depth minus seat thickness minus working length should leave 1–2 mm clearance. If the math does not work, re-measure before you order — not after the seal arrives.
Take a photo of the measurement setup and the old seal before you discard anything. More than once, a supplier has called back with a question that only the original hardware could answer. Keep the old seal until the new one is installed and running.