Laboratory beam test, twelve specimens, three mixes

Civil engineering research poster example

A structures poster has to make a reader trust twelve specimens, which means the casting, the rejections and the failure mode all belong on the board. This one runs seven narrow columns across A0 landscape, from abstract to references, with the specimen flow and the load results in the middle where a reviewer will look first.

Create a poster with OneCraftA0 Landscape, printed at 1189 by 841 millimetres

The whole board

The poster at full size, exactly as it prints. Every number, citation and caption on it was written for this example, so the layout is being judged on real content.

MU
Recycled aggregate concrete beams in four point bending: twelve beams, three mixes, one control
S. Ravenhill ¹, F. Adeyemi ¹, G. Lindsay ²
1 Structures Laboratory, Marrandah University · 2 School of Civil Engineering, Marrandah University
Abstract
Recycled coarse aggregate is available, cheap and under used, and the usual objection is that nobody trusts it in a structural member. Twelve reinforced beams were cast in three mixes, replacing 0%, 30% and 50% of the coarse aggregate by mass, and tested to failure in four point bending. Mean ultimate load was 142.6, 137.4 and 124.8 kN. The 30% mix sits 3.6% below the control, which is inside normal batch scatter, and the 50% mix sits 12.5% below it.
Introduction
Australian demolition waste is crushed and graded in volume, and most of it goes into road base rather than into structural concrete, because the design codes that would let it into a beam ask for evidence the industry has not produced at a useful scale. The two objections raised most often are strength loss and serviceability: a recycled mix carries old mortar on the aggregate face, which lowers stiffness and widens cracks at working load even when the failure load looks acceptable. This programme tests both objections on the same specimens. It is deliberately small, twelve beams of one section and one reinforcement ratio, so it answers a narrow question well rather than a broad one badly. The question is whether a 30% replacement behaves like the control closely enough that a designer could treat it as the control.
METHODS
Specimens. 12 beams at 200 by 350 mm over a 3,000 mm span, four per mix, all with 3 N16 bars bottom and N10 stirrups at 150 mm.
Mixes. Recycled coarse aggregate replacing 0%, 30% and 50% by mass. Cement content, water to cement ratio and sand were held constant.
Curing. Moist cured 7 days, then stored at 23 degrees and 50% humidity. All beams tested at 28 days, within a four day window.
Loading. Four point bending, 1,000 mm constant moment zone, displacement controlled at 0.5 mm a minute to failure.
Measures. First crack load, crack width at 0.6 of ultimate, midspan deflection and ultimate load, from load cell and two LVDTs.
PARTICIPANT FLOW
Beams cast across the three mixes
15
Rejected, honeycombing or cover depth out of tolerance
3
Beams tested in four point bending
12
Control mix, no recycled aggregate
4
Recycled mixes at 30% and 50%
8
First crack
Ultimate
0%
30%
50%
04080120160
Figure 1. First crack (lower bar) and ultimate load (upper) for the 0%, 30% and 50% recycled mixes, mean of four beams, in kilonewtons.
Source: Marrandah Structures Laboratory, 2026.
Results
Ultimate load, crack width at service load and midspan deflection at ultimate, by mix and as the mean of four beams: the control reached 142.6 kN, 0.18 mm and 24.6 mm; the 30% mix reached 137.4 kN, 0.21 mm and 26.1 mm; the 50% mix reached 124.8 kN, 0.29 mm and 31.4 mm. Within mix scatter on ultimate load was 4.1, 5.3 and 6.8 kN, so the 30% result sits inside the control scatter and the 50% result does not. Every beam failed in flexure by yielding of the tension steel followed by crushing of the compression face, with no shear or anchorage failure in any specimen. First crack arrived at 48.2, 45.6 and 39.1 kN, which is 33.8%, 33.2% and 31.3% of each mix ultimate load, so the recycled mixes crack slightly earlier in proportion as well as in absolute terms. Stiffness in the uncracked range, taken as the secant to first crack, fell 6% at 30% replacement and 14% at 50%, which is the single number that explains the crack width and deflection results below.
Discussion
Strength is not where recycled aggregate costs you here. Serviceability is. The 30% mix lost 3.6% of ultimate load, which a designer would not notice, but its crack width at working load rose 17% and its deflection rose 6%, both driven by the lower stiffness of the old mortar on the aggregate face. At 50% the same mechanism doubles: 12.5% off the ultimate load and a 61% wider crack. A serviceability governed member would therefore reach its limit at 50% well before a strength check noticed anything. The practical consequence for a designer is that a replacement ratio cannot be signed off on a cylinder test. A 50% mix that meets its specified compressive strength will still deflect and crack more than the control, because the loss sits in stiffness rather than in strength, and stiffness is not what a cylinder reports.
Future work
Four beams a mix is enough to see a mean and not enough to set a characteristic value, so the next programme is twelve beams a mix at 30% only, with three source yards rather than one, to put a number on the variation that source introduces. Long term deflection under sustained load is the second gap, because every measurement here is short term and the creep of old mortar is exactly where a recycled mix would be expected to behave differently. A third question worth answering is whether a higher reinforcement ratio recovers the serviceability loss at 50% cheaply enough to be worth specifying.
3.6%
below the control at 30% replacement
137.4 kN against 142.6 kN, mean of four beams each
Key finding
At 30% recycled coarse aggregate the beams carried 137.4 kN against 142.6 kN for the control, a loss of 3.6%, inside batch scatter.
30% mix within 3.6% of the control on ultimate load
50% mix 12.5% below the control, outside batch scatter
Crack width at service load 0.18, 0.21 and 0.29 mm
All twelve beams failed in flexure, none in shear
REFERENCES
1.Gillard, N. and Prabhu, V. (2025) Recycled coarse aggregate in flexural members: a review of 41 test programmes. Australian Structural Engineering Journal, 22(1), 5 to 31.
2.Teoh, W. (2024) Serviceability rather than strength: the real limit on replacement ratio. Concrete Research Letters, 15(3), 144 to 160.
3.Adeyemi, F. and Ravenhill, S. (2026) Batch scatter in twelve beam programmes. Marrandah Structures Reports, 9(2), 70 to 88.
4.Lindsay, G. (2023) Source yard variation in crushed demolition aggregate. Journal of Construction Materials, 18(4), 255 to 273.
5.Marrandah University Structures Laboratory (2026) Four point bending protocol for reinforced beams, version 6.

Block by block

What each block on the board is for, in the order a reader walks it.

Title band and three authors
The title gives the test, the specimen count and the fact that one mix is a control, which sets up the three way comparison the columns underneath carry.
Abstract
Why recycled coarse aggregate is under used in structural members, the twelve beams and three mixes, and the three mean ultimate loads with the 3.6% and 12.5% losses stated.
Introduction
The two objections raised against recycled aggregate, why old mortar on the aggregate face lowers stiffness, and why a programme of twelve beams answers a narrow question well rather than a broad one badly.
Methods as a labelled protocol
Five entries: the beam geometry and reinforcement, the three mixes with cement and water held constant, the curing regime, the displacement controlled four point loading, and the five measures taken.
Specimen flow
Fifteen beams cast, three rejected for honeycombing or cover depth outside tolerance, twelve tested, four in the control and eight in the recycled mixes.
Figure 1, first crack and ultimate load
Two bars per mix on the same kilonewton scale, mean of four beams each. This figure block has its legend switched off, so the caption names the lower and upper bars.
Results and the headline number
Ultimate load, service crack width and midspan deflection for all three mixes, the within mix scatter, the flexural failure mode in every specimen, and the stat block carrying the 3.6% loss at 30%.
Discussion, future work, key message and references
Why serviceability rather than strength sets the usable ratio, what the next programme should test, the callout carrying the two ultimate loads, four checked code implications and five numbered references.

What makes this board work

The rejected specimens are on the board

Fifteen beams were cast and three were rejected for honeycombing or cover depth outside tolerance. Putting that in the flow block, rather than reporting twelve as if twelve were cast, is what lets a reviewer believe the twelve that were tested.

Two loads per mix in one figure

First crack and ultimate load are plotted together for all three mixes, which shows that the recycled mixes crack earlier in proportion as well as in absolute terms. Three categories and two series is the largest comparison this narrow column holds cleanly.

The conclusion contradicts the obvious reading

Ultimate load barely moves at 30%, so the board spends the discussion on serviceability instead: a 17% wider service crack and a 6% lower stiffness. Saying that a cylinder test cannot sign off a replacement ratio is the finding a designer needs.

Questions people ask

What goes on a civil engineering research poster?

The specimen geometry and reinforcement, the mix or material variable, the loading arrangement and rate, the instruments, the results with their scatter, and the failure mode of every specimen. A structures poster that omits the failure mode leaves a reviewer unable to judge whether the test measured what it claims.

Should I show rejected specimens on a poster?

Yes, and in a flow block if the design has one. This board shows fifteen beams cast, three rejected with the reasons given, and twelve tested. A reader who sees only the twelve has no way to judge selection, and three rejections out of fifteen is a normal casting yield rather than a problem.

How do I report scatter with only four specimens?

Give the range rather than an interval that four points cannot support. This board reports within mix scatter on ultimate load of 4.1, 5.3 and 6.8 kilonewtons and uses it to argue that the 30% result sits inside control scatter while the 50% result does not.

Do Australian structural posters need local context?

When the point is a code change, yes. This board is written around Australian demolition waste that is already crushed and graded for road base, and the code implications list is what a local reviewer will read first. State the jurisdiction so the implications are not read as universal.

Which poster layouts suit a seven column board?

A broadsheet layout like this one, where narrow columns each hold one block, suits a study with many short sections and a small figure. It is the wrong choice for a study whose figure is the finding, because the figure column here is only 432 pixels wide on the working canvas.

What does an A0 landscape poster export at?

The working canvas is 2,340 by 1,656 pixels and the export captures at twice that, so 4,680 by 3,312 pixels, roughly 100 dots per inch at full A0 size. It is RGB and carries no bleed, no crop marks and no CMYK conversion, so a print shop needing those will prepare them.

Build your own in about a minute

The button below opens the generator with this use case already described. Change the wording to match your own, generate, then edit anything you like.

Make one like this

Other poster examples

Architecture research poster example

An architecture poster that reports a measured building has to open with the question, because the reader has to know what was being compared before the photographs they expect fail to appear. This board puts the question in a title panel down the left, then walks 24 classrooms, three window designs and twelve months of readings across the remaining columns.

Software engineering research poster example

A software engineering research poster has to make a mined dataset readable from two metres away. This A0 landscape board does it for code review: 41,000 pull requests in 14 open source repositories, with the median wait for a first review falling from 31 hours to 22 over three years, shown as a four point timeline across the top.

Data science research poster example

A model comparison poster is read as a table first, because the ranking is the finding. This board puts five scorers in one table, the rules list currently in use and four models, all measured on the same held out set of 25,200 subscribers, and leads with precision in the top decile rather than with accuracy.

Want the steps in the builder? Read Add charts and diagrams, then choose the template, theme and size. For everything this generator can do, see the poster maker.

Sources

Written and checked by the OneCraft team. Last checked .