International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 09 Issue: 07 | July 2022 www.irjet.net p-ISSN: 2395-0072

Experimental and Computational Fluid Dynamics (CFD) Analysis of

Additively Manufactured Weirs
Dr. Fayyaz Rehman1, Robert Benham2
1AssociateProfessor, Faculty of Creative Industries, Architecture and Engineering, Solent University, Southampton,
UK
2Senior Lecturer, Faculty of Creative Industries, Architecture and Engineering, Solent University, Southampton, UK

---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Additive Manufacturing is emerging as a cost- stage to ensure the manufactured parts conform the quality
effective alternative to conventional manufacturing requirements. Additive manufacturing (AM) printed parts
techniques for applications requiring components with offer a number of distinct advantages over conventionally
complex geometries, assemblies comprising a large number of machined components, in terms of production time, cost, and
parts or small productions runs. Cost savings can be realized the possibilities of achieving complex geometries. Weir
through reduction in raw material required, reduced improvements in their designs have historically required a
manufacture times and removing the need for expensive substantial amount of empirical testing. The existing
tooling. AM can offer an economical alternative to the existing experiment used at the University for undergraduate
alloy weir design to perform fluid mechanics experiment in our students studying Mechanical Engineering uses a 2.5 m long
lab. An existing 2.5 m open channel fluid flow experiment flow channel (Figure 1) which allows for many different
contains a set of standard weirs which is limited to sharp experiments to observe the behaviour of open channel flow
crested flat profile in design. This paper will compare with various components. The existing apparatus is supplied
experimental AM weirs (e.g., labyrinth, piano, catenary), that with machined parts, including simple weirs which are
would not be possible on some laser-cut polymer or machined limited to one sharp crested flat weir design (Figure 2),
aluminum weirs. Due to the bespoke complex nature of weirs’ sluice gates and a flat Venturi channel. Most parts are
design other manufacturing methods would be too expensive machined; some polymeric parts are injection moulded. The
and impossible to use. AM technology allows a cost-effective flow channel can be used to test more complex geometry
solution for progressive design modifications to be parts. Multiple weirs can be created using AM technology
implemented throughout investigations. This paper will with differing geometries at a much lower cost than the
highlight comparisons made between a range of AM produced machined alloy parts. Leadtime is around 6 - 8 hours for the
weirs in terms of flow rate, fluid velocity profile, water level components in this study. During the experiment students
height and discharge coefficient. Computation fluid dynamic record a range of volumetric flow rates and determine key
modelling (CFD) will also be used to verify, analyze, and fluid parameters. In addition, computational fluid dynamic
compare results. Based on the experimental results and (CFD) modelling can be used to simulate the flow and predict
verification, the paper will also discuss the suitability of parameters that cannot be estimated accurately from the
application of AM techniques in fluid flow analysis experiment. Therefore, computation fluid dynamic modelling
experiments. (CFD) will also be used to verify, analyse, and compare
results. Based on the experimental results and verification,
Key Words: Additive Manufacturing, Photopolymer Resin, the paper will also discuss the suitability of application of
Experimental Methods, CFD Analysis AM techniques in weir design and analysis.

1. INTRODUCTION
Additive Manufacturing is relatively a new material addition
technology to design and manufacture production ready
polymeric and metallic components as compared to the
classical manufacturing processes such as machining, casting
or moulding. They allow new innovative design to be
produced with regards to material, shape and complexity of
the part because these manufacturing processes eliminate
the need of tooling.
A lot of current restrictions of design for manufacturing and
assembly are removed due to the use of these AM processes. Figure 1: A 2.5 m long flow channel along with hydraulic
However, AM processes have their own characteristics and tank
requirements which need to be considered during the design

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 07 | July 2022 www.irjet.net p-ISSN: 2395-0072

Figure 4: Labyrinth and Piano Weirs

Figure 2: Existing sharp crested flat weir design

2. RESEARCH METHDOLOGY
Due to the limited time and preliminary testing, the number
of weirs was restricted to four design models that could be Figure 5: AM weirs (from left to right: empty catenary,
compared to the supplied machined sharp-crested weir. The ramped catenary, labyrinth, piano key).
weirs selected for this study include two with a catenary
crest (one empty and one ramped in Figure 3), a labyrinth The procedure for performing the experiment is outlined
weir and a piano key type weir (Figure 4). New weirs below:
(Figure 5) are made from tough additive manufacturing  The flow channel is connected to the hydraulic tank and
polymer, like ABS, a translucent plastic type commercially an angle of 0.5° (entry level higher than exit level). The
available photopolymer resin using a Stereolithography internal dimensions of the channel were recorded (53
(SLA) AM machine. mm wide x 120 mm deep).
 The pump is switched on and set to maximum flow.
 The height of the water level is recorded at designated
positions along the channel, 200 mm apart. These
positions remain constant once the weirs are inserted.
 A pitot tube connected to a data logger is used to
measure the velocity of the fluid over 20-second
intervals. Statistics tools give the mean, maximum and
minimum values. The process is repeated twice to
check consistency of readings.
 The volumetric flow rate was recorded by timing the
tank to fill to 35 litre capacity. From the data, the
velocity is calculated (and compared), the Reynolds
number, Froude number as well as flow rates.
 Each weir independently is then inserted 1800 mm
from the entry end, and stages 2 – 5 repeated.
 The experiment procedure is repeated for minimal flow
Figure 3: Empty and Ranked Catenary Weirs (ensuring water coverage on the walls of the weir).
Two flow rates were used: maximum flow (to mimic flood
Each weir had a flat crest; further research will look at
conditions) and minimum flow (ensuring full coverage of the
variations in crest profile. The flow channel was run empty
weir wall). The datum for taking height measurements was
and recorded when set at 0.5° at full flow. A number of
the top of the flow channel and depth to the water level was
parameters were calculated, such as volumetric flow rate,
measured by a depth gauge. Points were measured at 200
average velocity, the Reynolds number, Froude number. The
mm intervals close to the weir upstream and 600 mm from
weirs were then tested at maximum flow rate and then a
the weir downstream. A pitot tube connected to a data logger
reduced flow rate so that the range in upstream height could
was used to measure the velocity and a representative value
be calculated and the weirs could be compared under
was recorded at each of the measuring points using the
different flow conditions. Computational Fluid Dynamic
statistics tool over 20-second time periods.
(CFD) modelling based on flow simulation was carried out

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 07 | July 2022 www.irjet.net p-ISSN: 2395-0072

using the actual flow rates, so that the variations in velocity

profile could be seen and regions of interest established.
A spreadsheet was used to compute the flow rates, Froude
and Reynolds numbers as well as carry out comparisons to
the supplied (machined) sharp-crested weir. The main
comparison factors used are the average height of the water
marks upstream; the differences in water heights between
the maximum and minimum flow rates; the discharge
enhancement ratio, referred to here as the “r” value (see
equation 1); the mass-flow rate ratio of discharge between
the upstream and downstream and the mass flow rate
downstream. CFD modelling was used to compare the
calculated the parameters and in addition observe the 3.2 Testing of Weirs
velocity profiles at regions in the channel such as the weir One factor that limits the testing of weirs is the relatively
approach, the crest, nappe, and other sections of the channel narrow width of the available flow channel (53 mm). A lot of
far away from the weir. research employs wider channels, although the results
overall show that the AM weirs had good consistency for the
A spreadsheet (Table 1) was used to compute important upstream water height a maximum flow. One problem with
results parameters for different flow rates. geometry of the weirs coupled with the parameters was to
try and determine the coefficient of discharge accurately. As
Table 1: Data Collection for different flow rates mentioned by several authors, notably Kumar et al (2020)
the determination of these parameters is complicated often
requires different modelling approaches. Unlike the sharp-
crested weir that has a relatively straight forward formula to
calculate the coefficient of discharge, Cd, the other weirs
could be compared instead by the discharge enhancement
ratio:
3

QAM weir Cd Leffective 2 g H 2 Leffective

r  3
 (1)
QSharp crested weir 2
W
Cd W 2 g H

3.3. Comparison of Maximum and Minimum Flow Rates

Initial testing of the weirs showed very clearly that variation
3. ANALYSIS OF RESULTS in upstream height between the maximum and minimum
flow rates was much lower than the sharp-crested weir. For
3.1 Weir Theory maximum flow rate, the simple catenary weir exhibits the
For a sharp crested weir (Figure 6), the computation of best combined properties. The comparisons (Table 2) are
discharge is calculated relatively simple. The labyrinth weir scored the lowest in all
parameters apart from average height of water upstream.

Table 2: Maximum Flow Rate Comparisons for Weirs

Figure 6: Sharp Crested Flat Weir Flow Parameters

For minimum (reduced) flow rate, again the catenary

remains the better model however the comparisons (Table

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 07 | July 2022 www.irjet.net p-ISSN: 2395-0072

3) are more complicated. The piano key weir does not channels. Labyrinth weirs are attracted a lot of interest due
perform in majority of parameters such as average height of to their relatively long existence and widespread use; studies
water upstream, Coefficient of discharge and maximum flow by Crookston, Paxson and Campbell (2013) do compare
rate. several geometrical factors and include a 2-cycle labyrinth
weir to others.
Table 3: Minimum (Reduced) Flow Rate Comparisons for
Weirs

Figure 9: Side-on view of the labyrinth weir at full flow

rate.
Another aspect that can be easily compared is the gradient of
the water line over the side length of the weir, and the A logical next step in this research would be compare
difference in forms of nappes at the flow rates for both different crest profiles (e.g., sharp, flat, quarter-round, half-
empty (Figure 7) and ramped (Figure 8) catenary type of round, ogee). The catenary shape derived from the
weirs at full flow rate. hyperbolic cosine function (cosh x) was chosen by authors
due to the relatively long crest that can be achieved in a
relatively narrow channel.
One issue that could be improved in terms of design for
testing AM parts is reducing the side wall and under some
weirs, this was most noticeable with the piano key weir
(Figure 10). The most effective weir preventing this was the
ramped catenary weir, due to the increased rigidity of the
model.

Figure 7: Side-on view of the empty catenary weir at full

flow rate.

Figure 10: Side-on view of the piano key weir at full flow
rate.
The piano key weir (figure 10) design was based on similar
design ratio parameters from other research (Erpicum et al),
and at high flow rates scored well. The most consistent in
Figure 8: Side-on view of the ramped catenary weir at full terms of ranking over all comparators is the ramped
flow rate. catenary, in third place at maximum flow but second at
lowest flow. The error of the pitot tube was evaluated in the
The labyrinth (Figure 9) weir did not compare well in on empty channel, where velocity was calculated from the
most of the comparison criteria for both flow rates, however volumetric flow rate. It was found to be 96% accurate, from
the major limitation of only having two cycles meant that the readings taken.
weir characteristics are unlikely to be similar for much wider

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 4
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 07 | July 2022 www.irjet.net p-ISSN: 2395-0072

3.4 Experimental Issues.

Although results show consistency, however following
experimental issues were encountered:
 The channel is relatively narrow, limiting the number of
repeating sections of the weir.
 The ramped catenary weir had the optimum rigidity and
required the least rubber mount support. The piano key
was the least stable.
 Side leakage and floor leakage were observed on all
weirs; the piano key showed considerable side leakage.
 The pitot tube position had to be securely fixed,
otherwise considerable variation was probable.
 Limited time only allowed for a few weirs to be tested
for this research.
 Only flat crest profiles were tested for the initial Figure 12: Flow simulation of ramped catenary weir at
research. This profile is not optimum. maximum flow rate
 The coefficient of discharge is easy to calculate for the
sharp-crested weir but more complicated for other
weirs. The narrow channel problem means that the
coefficient is likely to be very different for a much wider
channel with more repeating sections.
 As an alternative the “r” value was measured (Q test
weir/Q sharp-crested weir).

3.5 Finite Element Analysis Flow Simulation

Finite Element Analysis based CFD flow simulations were
done to evaluate and compare the flow dynamics results
obtained from experimental data for all four different types
of weirs. The fluid flow pattern observed through CFD
analysis of empty (Figure 11) catenary, ramped (Figure 12)
Figure 13: Flow simulation of labyrinth weir at maximum
catenary, labyrinth (Figure 13) and piano key (Figure 14)
flow rate
weirs is almost like experimental flow pattern.

Figure 14: Flow simulation of piano key weir at maximum

flow rate

Figure 11: Flow simulation of empty catenary weir at Simulation results also verified and compared the average
maximum flow rate velocities obtained from experimental data for both empty
and ramped catenary weirs shown in Table 4. Change in
shape from empty to ramped shaped weir results in
negligible reduction in average velocities from CFD analysis.
Simulation analysis was also very useful for regions that are
difficult to measure with the pitot tube and boundaries
within the system.

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 07 | July 2022 www.irjet.net p-ISSN: 2395-0072

Table 4: Comparison of Experimental/Simulation Data for  The advantages of AM are clear in this study: huge
average velocities in both catenary weirs. reductions in lead-times and very economical parts
can be produced.
Type of Catenary Weir Experimental Average
Velocity (m/sec)
CFD Average Velocity
(m/sec)
 Results are consistent and repeatability was good.
Empty 0.650 0.628  A limited number of weirs were tested. There is a
Ramped 0.650 0.627 huge scope of weirs that could be experimented
with, but the next step is to vary the crest profile
Overall results yield following key points:
(e.g., rounded, ogee).
 The scope for further testing is great. The results in
 To achieve optimization, weirs in wider test flow
this small study show the potential of a catenary
channels should be tested. AM models can be made
shaped weirs, however other characteristics need to
in sections and slotted together to accommodate
be considered, such as coefficient of discharge.
larger widths.
 AM weirs are lighter than the supplied metallic ones
 The next stages are to look at flow induced design
and do need supporting mounts. The side and
optimisation
bottom leakage could not be measured. Reducing
this is important before further work can
commence. REFERENCES
 AM parts took around 6 - 8 hours to produce and
[1] FormLabs tough and durable material, Formlabs
cost around £12 each. By introducing complex material, 2022,
profiles, this does not alter these values very much. https://formlabs.com/uk/materials/tough-durable/ ,
Several parts can be produced in one period. [Assessed on 01 June, 2022].
 Some AM profiles could not be produced by [2] Kumar, M., Sihag, P., Tiwari, N. K., & Ranjan, S. (2020).
conventional methods on aluminum alloys or Experimental study and modelling discharge coefficient
Perspex models. The tough AM SLA resin is well of trapezoidal and rectangular piano key weirs. Applied
suited to this application. Water Science, 10(1), 1-9.
 Durability for some AM polymers could be an issue, [3] Crookston, B., Paxson, G., & Campbell, D. (2013).
however none of those tested raised any concerns Effective spillways: Harmonizing labyrinth weir
hydraulic efficiency and project requirements. Labyrinth
and these will be used again. and Piano Key Weirs II, 283.
[4] Erpicum et al, Labyrinth and Piano Key Weirs. (2011).
4. CONCLUSIONS United Kingdom: Taylor & Francis.
[5] Bilhan, O., Emiroglu, M. E., Miller, C. J., & Ulas, M. (2018).
The research shows that effective weir designs can be The evaluation of the effect of nappe breakers on the
implemented and compared using suitable AM technologies. discharge capacity of trapezoidal labyrinth weirs by
The tough AM SLA resin performs well although the ELM and SVR approaches. Flow Measurement and
flexibility in some designs requires further structural rigidity Instrumentation, 64, 71-82.
around the base to prevent the deformation of the shape
when held in the channel. The catenary weirs show BIOGRAPHIES
promising results however the regions of interference need “Dr. Fayyaz Rehman is an
to be studied further, and extra features could be added to Associate Professor at Warsash
try and reduce the detrimental factors on performance. For School of Maritime Science and
example, nappe breakers have been experimented with in Engineering, Solent University, UK.
several studies, Bilan et al (2018), and AM can allow for He is a Fellow of Higher Education
more complex geometries to be investigated. Further Academy, a Chartered Engineer
investigation is required prior to this to compare different from the Engineering Council and a
forms of crest, as the flat crest is not optimal. In conjunction Fellow of the Institution of
with this, coefficients of discharge need to be mathematically Engineering Designers, UK. He is
determined and compared to the measured flow rates over also vice chair and committee
the weir. The volumetric flow rates can then be compared member of the Consortium of UK
against the changes in height upstream to give a better Manufacturing Engineering Heads
comparison. The relatively narrow channel does limit the (COMEH), a UK-based body
amount of work that can be carried out; however, the results responsible for promoting
show that even with flat crested weirs the variation in weir manufacturing engineering
design can be effectively compared by using AM technology, education and research, as well as
which can give a good advantage over other construction organizing the International
methods of model weirs for open channel flow studies. Conference on Manufacturing
Overall, from this study, we can conclude that Research (ICMR) conference series

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 6
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 07 | July 2022 www.irjet.net p-ISSN: 2395-0072

annually. His research interests are

CAD/CAM/CAE, Material Testing
and Additive Manufacturing
Technologies.”

“Rob Benham is a senior lecturer

in engineering science and
currently course leader for the
HNC Engineering and Foundation
Year Engineering. He has been
teaching on engineering courses at
Solent University for over 15 years.
Prior to the introduction of
individual course leaders, he was
programme leader on the old
engineering programme. In 2003
Rob completed his PGCE (post-
compulsory education) at Oxford
Brookes University. He then taught
in further education for one
additional year. He continued some
FE teaching and supply teaching
when working part-time. In
addition to this, in more recent
times, Rob has delivered many
taster days at Solent. He has strong
research interests in
manufacturing and materials. As a
result, he brings a wide range of
experience of educational settings
with a diverse scope of learners.”

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 7