Module 7 · Lesson 7.1
Bar arrangement and effective depth
Where a design quietly loses capacity that no check will ever report.
Why this matters
A beam is designed for 2100 mm² of steel, and 5 H25 is specified. On site the fixer finds they will not go in one layer, so they go in two — three and two. Nobody tells the designer, because nothing has gone wrong from the fixer's point of view. But the effective depth is now 25 mm less than the calculation assumed, the lever arm has shrunk, and the beam is about 5% weaker than the drawing claims. No check anywhere in the design process reports this.
By the end of this lesson you should be able to
- Compute the effective depth to the centroid of a multi-layer bar group
- Quantify what an extra layer costs
- Explain why the loss is invisible to a strength check
- Choose between more bars, bigger bars, and a wider section
What you should already know
- Flexural design and the lever arm (Module 4)
- Bar selection and fit (Module 8)
- Bond and anchorage (Module 5)
Effective depth is a property of the built thing
Every flexural calculation in this course has used d, the effective depth to the centroid of the tension reinforcement. In a single layer that is simply
d = h − cover − link diameter − half the bar diameter
Put the bars in two layers and the centroid drops. For a group of n₁ bars at depth y₁ and n₂ at y₂:
ȳ = (n₁y₁ + n₂y₂)/(n₁ + n₂), and d = h − ȳ
That is all there is to it arithmetically. What matters is that the design must use this value, and that the value is not known until the bar arrangement is known — which is after the design is done.
So the sequence has a loop in it: design assuming one layer, choose bars, discover they need two, recompute d, redesign. Module 8 called this the circular dependency and broke it with one revision. This is the revision.
Try it
Lay the bars out
Add bars until they spill into a second layer, and watch the effective depth fall. No strength check reports that loss — it simply happens.
- Layers needed
- 1
- Bars per layer
- 4
- Effective depth d
- 544.5 mm
- d if one layer were assumed
- 544.5 mm
- Effective depth LOST
- 0.0 mm
- Steel area As
- 1963 mm²
- Steel ratio
- 1.09%
- Width occupancy
- 47%
- Buildability
- comfortable
- Crack control
- by spacing
The bars sit comfortably and concrete will flow around them.
Things worth trying
- Add bars one at a time. There is a moment when the count exceeds what one layer holds, and d drops by 25 to 40 mm at a stroke — with no warning from any strength check.
- Widen the beam by 50 mm at that point. The bars fold back into one layer and the depth returns. Width is cheap; depth is not.
- Raise the cover from 35 to 50 mm. It costs effective depth AND makes crack control harder, because crack spacing grows with cover.
- Raise σs towards 360. Crack control tightens sharply, and the fix is smaller bars more closely spaced — not more steel.
Worked example
What a second layer costs
Given
- Beam 300 mm wide × 600 mm deep, 35 mm cover, H8 links
- Design requires 2900 mm² of tension steel
- Candidate: 6 H25 = 2945 mm²
Find
Whether the bars fit in one layer, and what the arrangement costs.
Assumptions
- Minimum clear spacing is the greatest of the bar diameter, aggregate size plus 5 mm, and 20 mm
- Clear vertical gap between layers equal to the bar diameter
Predict first
A beam needs 6 H25 in a 300 mm web. The designer widens the beam to 350 mm. What happens to the effective depth?
Practice
A 300 mm wide beam has 35 mm cover and H8 links. How many H20 bars fit in one layer? Take the minimum clear spacing as 25 mm.
Practice
Bars are arranged 4 in the bottom layer at 55.5 mm from the tension face and 2 in a second layer at 105.5 mm. In a 600 mm deep beam, what is the effective depth, in mm?
Summary
- d is measured to the CENTROID of the bar group, not the bottom layer
- A second layer typically costs 15 to 40 mm of effective depth
- No strength check reports the loss; only the deflection check would, if repeated
- In a narrow web, fewer and larger bars usually beat more and smaller
- Widening the beam buys effective depth indirectly, by changing the layer count
This is educational material. It uses simplified examples to teach principles, and must not be relied on for real design or safety-critical decisions. Module overview and checkpoint