A packet that is taller at one end does not necessarily contain more steel there. Its height includes the steel, coatings and spaces between sheets. Uneven support or pressure can change those spaces while the sheet count stays constant.
Before removing a sheet or changing the build schedule, determine which component of height changes along the packet. Count and thickness describe material quantity. A matched-pressure height comparison probes the contribution of interfaces and restraint.
Write the height balance explicitly
At a location x, a simplified stack height is H(x) = Σtᵢ(x) + C(x) + G(x). The steel thicknesses tᵢ, total coating contribution C and intersheet gap contribution G are all in millimetres. This is a bookkeeping relation rather than a constitutive compression law.
Compare locations using the same measurement definition. A height measured over an edge peak cannot be directly compared with one measured between flat contact faces. The support datum and contact area must remain consistent.
| Possible cause | Expected evidence | Useful discriminating check |
|---|---|---|
| Different sheet count | Count or packet-build discontinuity | Trace layer identities and count at the affected region |
| Thickness variation across sheets | Material sum varies despite equal count | Compare representative thickness positions and coil records |
| Different interface gaps or compression | Height changes when matched pressure is applied | Repeat under controlled, equivalent contact conditions |
An apparent count change may instead be a staggered edge that makes layers difficult to see. Verify count with build records or an accessible reference region. Photographs of one obscured edge are weak evidence of missing steel.
Compare pressure response, not just final height
If approved for the assembly, measure both locations under equivalent support, contact area and applied load. Pressure equality is not established by using the same total force on different areas. Record the loading and unloading sequence because interface seating can depend on previous compression.
If the initial taper decreases markedly under matched conditions, interface closure is implicated. If it remains while counted steel thickness differs, the material build is implicated. If neither explains the pattern, investigate the datum, local waviness and clamp geometry.
Residual height after unloading deserves separate attention. A one-time seating movement differs from a reversible elastic response. Do not use a later compressed reading to claim that the original free-state geometry was correct.
Mixed causes are possible. A thickness gradient and an uneven clamp can reinforce each other or partly cancel. Keep the material and pressure-response evidence separate instead of assigning the whole height difference to the first visible clue.
Changing sheet count affects magnetic section as well as geometry. It should follow a confirmed build discrepancy and the approved design, rather than serve as a convenient way to flatten the packet. When the gradient comes from gaps, correcting seating or restraint addresses the cause while preserving the intended steel quantity.
Separate steel quantity from a measured taper
For 23ZH90, nominal thickness is 0.23 mm and catalogue thickness tolerance is ±0.02 mm [1]. Assume 100 sheets at both ends, initial heights of 24.0 and 24.6 mm, and matched-load heights of 23.8 and 24.0 mm. Nominal steel sum is 23.0 mm; the initial taper is 0.6 mm and the loaded taper 0.2 mm. A 0.4 mm reduction implicates interface seating, while the residual remains unresolved. The extreme arithmetic stock envelope is 100 × (0.23 ± 0.02) = 21–25 mm, not a statistical prediction. Obtain actual thickness and count records before removing sheets; include correlated material variation in the uncertainty model [2].
Manufacturing controls. Thickness distribution, layer count, coating, waviness, support flatness and clamp-pressure distribution control the gradient. A correlated coil thickness bias can accumulate; dividing its uncertainty by the square root of the sheet count would conceal that effect.
References
- Nippon Steel Corporation, Grain-Oriented Electrical Steel Sheets, catalogue D004je_03_202004, 2020, Specifications; Size and Form Tolerances.
- JCGM, JCGM 100:2008, Evaluation of measurement data — Guide to the expression of uncertainty in measurement, 2008.