Step-lap CRGO lamination design: how it reduces no-load loss and noise

Let’s talk about what actually changes in the core when you move to a well-designed step-lap CRGO lamination stack… and where engineers quietly lose the benefit on the shop floor.

1. What step-lap really changes in the core

At the joint, three things matter more than the brochure drawings admit:

• Local peak flux density at the corners
• Effective air-gap pattern through the stack
• How magnetostrictive forces add up in time and in space

Step-lap joints distribute the joints over several staggered overlaps instead of one single plane. Industry and academic work both show that, for a given steel grade and geometry, step-lap joints reduce no-load loss, excitation current, and sound pressure level compared with butt-lap or simple miter joints.

But that statement is almost useless without the details.

2. No-load loss: the real levers inside a step-lap pattern

You don’t reduce core loss “because step-lap is modern”. You reduce it because you control a small set of geometric and process variables.

2.1 Number of steps (3, 5, 7…)

• Multi-step vs single-step Studies comparing miter, single-step-lap and multi-step-lap joints on similar 3-phase cores show:
  • Step-lap designs with ~5 steps can cut total core loss by roughly 2–4.4% vs miter for the same CRGO and dimensions.
  • Moving from single-step to multi-step joints further improves loss and apparent power, but mostly at standard flux densities.
• Too few steps, too coarse Some noise-focused experiments indicate that 3-step patterns with small overlaps (≈2 mm) are poor from a noise perspective and don’t give consistent benefit. So the “cheap” 3-step pattern is often a half-measure in both loss and noise.
• Typical practical choice For distribution and small power transformers, 5-step is a workhorse. 7-step shows incremental improvement in loss at the cost of complexity and stacking effort.

2.2 Overlap length and step increment

Step-lap is essentially a controlled 3D air-gap pattern.

• Too short an overlap:
  • High local flux crowding at each step edge
  • Higher local loss and a bigger magnetizing current spike
• Too long an overlap:
  • Extra steel (cost)
  • More area where flux can wander between laminations

Design and test work on wound and stacked cores shows clear sensitivity of loss to lap length and number of laminations per step. There is usually a relatively flat optimum band rather than one magic value, and it shifts with lamination thickness and operating flux.

In practice, you’ll often see:

• Step increments: about 2–6 mm per step
• Effective lap length: tuned so the last step still closes cleanly without forcing, given your cutting and stacking tolerances

2.3 Flux density and “critical induction”

Multi-step-lap joints behave nicely up to a point – then not.

• Experimental work on 3-phase cores with multi-step-lap joints shows a critical induction: beyond this, apparent power and core loss rise faster and the multi-step pattern can even lose its advantage vs simpler joints.

What this means in design language:

• Don’t assume “we can push B higher because we have step-lap”.
• Treat the joint region with an effective permeability lower than the limb in your models.
• Use measurement on at least one prototype to locate that practical knee for your specific stack, clamping system, and steel batch.

2.4 Single-sheet vs double-sheet “books”

You know the building-factor game:

• Single-sheet assembly (one lamination per width in a “book”) gives a better building factor, fewer microgaps, and therefore lower losses.
• Double-sheet assembly makes handling easier but tends to increase no-load loss slightly, all else equal.

For a lamination supplier, this is where value quietly leaks: every time assembly switches from single-sheet to double-sheet without the designer adjusting the stack calculation, real no-load loss drifts away from the drawing.

3. Audible noise: why step-lap helps, and when it doesn’t

The noise story is mostly magnetostriction and how the joint geometry modulates it.

3.1 Typical benefit band

Field data and lab measurements agree on a rough range:

• Correctly cut and stacked step-lap CRGO cores often show 3–6 dB lower core noise vs similar non-step-lap cores at the same induction.

At low and medium inductions, multi-step-lap joints clearly reduce noise compared with miter or simple overlap. At higher inductions, the improvement shrinks and can flatten out, as some tests on model cores have shown.

3.2 Joint pattern and vibration spectrum

It’s not just “more steps = quieter”.

• Some 3-step patterns with small overlap lengths produce noise spectra that are not significantly better than non-step-lap designs.
• Multi-step-lap patterns spread magnetostrictive forces over a larger area and slightly shift the frequency content of the mechanical vibration – often away from the structural resonances of the tank and clamping frame.

So, if your acoustic problem is a narrow resonance in the tank, the right step pattern helps. If the problem is poor clamping or gaps, no geometry trick will save you.

3.3 Sensitivity to tolerances

Several industrial guides point out the same thing in different words:

The noise benefit of step-lap is strongly dependent on cutting accuracy, burr control, and assembly alignment. 

Misaligned steps, bent laminations, or uneven clamping reintroduce stress and small air gaps right where step-lap is trying to smooth the flux.

4. Design levers vs loss and noise – quick comparison

You can treat step-lap lamination design as a small parameter study instead of an art form.

5. Manufacturing reality: where step-lap gains disappear

On paper, step-lap is geometry. On the shop floor, it’s mostly sequence and discipline.

Key spots where lamination stacks decide your real loss and noise:

1. Cut-to-length and notching
   • Length tolerance directly impacts step positioning.
   • Notch burrs at the corners create microgaps exactly where flux is densest.
2. Guide holes and alignment pins
   • Multi-step-lap designs often use one or two guide holes per lamination to keep the stepping sequence correct.
   • If operators bypass pins “to save time”, the pattern drifts, and your measured loss looks like a different design.
3. Stacking order (“books”)
   • Theoretical stack calculations assume integer numbers of books per step. When the production team improvises because a book is damaged or missing, stack thickness and flux path change.
4. Stress-relief and flatness
   • CRGO lamination guides emphasize stress relief and flatness during annealing. Non-flat plates introduce bending stress when stacked, which hurts both loss and noise.
5. Core assembly and reassembly
   • Every time a core is opened and reassembled (factory test, transport, onsite inspection), step alignment can drift unless there is a clear procedure and marking system.

If you buy loose lamination stacks rather than finished cores, these points belong partly to your supplier and partly to your transformer factory. The interface is where problems usually show up.

6. How to specify step-lap lamination stacks in RFQs

If you want the benefits, you have to ask for them precisely.

Suggested items to spell out in an RFQ or technical specification for step-lap CRGO lamination stacks:

1. Steel grade and loss class
   • Nominal thickness and guaranteed core loss at specified B and frequency.
2. Joint concept
   • Step-lap with number of steps (e.g., 5 or 7).
   • Allowed joint types (no fallback to simple miter or butt-lap without written approval).
3. Geometric parameters
   • Target lap length and tolerance.
   • Step increment per step.
   • Maximum length tolerance for limbs and yokes.
4. Assembly method
   • Single-sheet or double-sheet books.
   • Required building factor or maximum stack height deviation vs theoretical.
5. Quality control on stacks
   • Sample core loss and excitation current test on assembled cores (yokes clamped) at specified B.
   • Visual criteria for burrs, coating defects, and damage at corners.
6. Noise expectations (if relevant in your market)
   • Even if you don’t specify a hard dB limit, you can request data from comparable step-lap designs showing measured sound pressure. Many suppliers already measure this.

This is how a lamination stack stops being a commodity and starts being a controllable part of your loss and noise budget.

7. A rough numeric feel: switching a 1 MVA core to step-lap

Take a 1 MVA, 3-phase, 3-limb stacked core at around 1.65 T in CRGO.

From published comparisons of miter vs 5-step-lap joints for similar cores:

• Total core loss drops by about 2–4% when moving from miter to 5-step-lap, holding steel and geometry constant.
• Apparent power (VA) drawn at no-load decreases more strongly (reported improvements on the order of 30% in some cases) because magnetizing current is sensitive to local saturation at joints.

For a core originally at 1600 W no-load loss:

• You might reasonably expect something like 1530–1560 W after a move to a well-executed 5-step-lap design, if all manufacturing conditions are under control.

On noise:

• If the original design was already mechanically decent, a 3–6 dB reduction in core noise is realistic, but only if your steel batch, cutting, stacking, and clamping meet the same standard as in the reference tests.

Treat those numbers as order-of-magnitude guidance, not guarantees. The actual spread between drawings and test reports usually comes from the manufacturing bullets in section 5.

8. Checklist before you sign off a step-lap lamination design

Use this list as a quick filter when reviewing drawings, offers, or factory proposals:

• [ ] Step count and pattern defined (e.g., 5-step single-sheet books) and documented on drawings
• [ ] Lap length and step increment specified with tolerances, not just nominal values
• [ ] Bmax in limbs checked against data for critical induction from similar cores or prototypes
• [ ] Building factor target included, with corresponding stack height limits
• [ ] Cutting, burr limits, and coating requirements written, not assumed
• [ ] Assembly method, guide holes, and stacking sequence defined in work instructions
• [ ] Acceptance tests for core loss, exciting current, and (where relevant) noise agreed with supplier

If any of these are missing, the phrase “step-lap CRGO lamination” on a quotation doesn’t tell you very much.

FAQ: Step-lap CRGO laminations, no-load loss, and noise