DESIGN AND TESTING OF AN ISO 5801 INLET
CHAMBER TEST RIG AND RELATED ISSUES WITH
Stefano CASTEGNARO1, Massimo MASI2,
1University of Padova, Department of Industrial Engineering,
via Venezia 1, 35131 Padova, Italy
2University of Padova, Department of Management and Engineering,
stradella San Nicola 3, 36100 Vicenza, Italy
A rig has been designed and constructed according to the ISO 5801 Standard, to test fans with a
maximum inlet diameter equal to 0.8 m. The rig design procedure, that turned into an effective
manufacturing and installation, is described. Issues that emerged from the preliminary study of the
5801 Standard are highlighted and discussed. Suggestions to improve the Standard are proposed.
This paper describes the design of an inlet-chamber fan test rig coupled with a multi-nozzle system
for flow-rate measurements, providing a procedure for the rig dimensioning and a discussion on some
issues and ambiguities encountered within the ISO 5801 Standard.
By April 2018 the new ISO 5801 Standard  shall be adopted as national norm in every European
country, superseding the previous 2008 version . The new Standard is the latest achievement of
a long and complex process of international standardization, which began in 1963 collecting recom-
mendations from several previously existing national fan testing codes [2, p. viii]. Likely because
of the complexity involved in this exercise of synthesis, some of the preceding versions of the Indus-
trial fans-Performance testing using standardized airways codes were not so simple to read and use.
Wallis, in fact, claims that preceding fan test codes came out from a “piecemeal process”. In past
years, the use of the Standard even required additional documents specifically addressed to the user
(e.g., ). The latest version of the Standard  results from a remarkable effort of simplification and
improvement. In spite of this enhancement, however, some issues still remain from the perspective
of a rig designer:
a) even if the requirements of all the components and rig parts are identified, a direct rig design
Darmstadt (Germany), 18-20 April 2018 2
procedure is not available in the Standard or in the literature;
b) some issues and incoherent informations exist within the Standard.
The second item, b), is particularly tedious for the rig designer, because issues and ambiguities translate
into a deeper analysis of the flow phenomena and the consequent interpretation of the Standard. This
increases the time required for the design process and, in turn, increases the costs.
In 2017 a rig was designed and constructed according to the ISO 5801:2008 code at the University of
Padova, to test fans with a maximum inlet diameter equal to 0.8 m. The rig features an inlet chamber
configuration and a multi-nozzle system for the flow-rate measurements. The facility was put into
operation in July 2017 and is currently operating both for research and didactic purposes.
In this work, parts of the technical and design achievements acquired from that experience are col-
lected. In the first part of the paper, the design procedure used for the rig is presented: the pro-
cedure resulted in an inexpensive design that was easy to assemble. The operative differences on
rig-chamber dimensioning between the new ISO 5801:2017 Standard and the previous 2008 version
are highlighted. In the second part, some issues and inconsistencies encountered within the Standard
are highlighted and discussed.
This paper is addressed toward fan testing rig designers and users. Due to the technical matter of
the discussion, previous knowledge of the Standard is required. In particular, all the symbols and
nomenclature are coherent with references [1, 2, 5].
The aim of the paper is twofold: i) provide a direct procedure that will result in a satisfactory design of
an inlet chamber test rig, and ii) opening a discussion on some issues contained within the ISO 5801
Standard and suggesting possible corrections/improvements.
Rig Configuration Selection
According to preliminary research performed on existing rigs, fan manufacturers appear to prefer the
chamber rig configuration (i.e., the A type ), instead of ducted ones (i.e., B,C,D layouts ). One
reason that makes this configuration valuable for the industries is that chamber rigs allow for testing
fans with widely different sizes (e.g., [6, p. 354]) without substantial rig modifications.
Ideally, the selection of the inlet or outlet chamber configuration provides no difference in terms of
assessing the fan performance. However, for the outlet-configuration layout the flow-field within the
chamber is complicated by the fact that the fan-exhaust-jet kinetic energy must be absorbed by the
flow-settling means upstream of the measurement planes and this results in some recirculation at the
jet boundaries.1However, similar conditions might also occur for an inlet-chamber rig that features a
multi-nozzle flow-measurement system, if a high-velocity jet is exhausted by some of the nozzles.
Both Inlet and Outlet test rigs were originally installed in the Laboratorio di Macchine Aerauliche e
Termiche at the University of Padova (see Fig. 1), with the first employed mainly with 315 mm axial
fans (e.g.. ) and the second with smaller cross-flow fans (e.g., ). The inlet chamber layout was
lastly selected in continuation with an on-going research on axial fans.
Design differences exist from the structural point of view, as walls are subject to a negative pressure
(i.e., they are squeezed toward the inside) in the inlet chamber case, while in the outlet chamber case
1The 2017 ISO Standard [1, p. 26] requires that tests are performed to verify that such recirculation is not excessive at
the fan static pressure measurement plane in the outlet chamber case.
Darmstadt (Germany), 18-20 April 2018 3
Figure 1: Original Inlet and Outlet Chamber Test rigs.
walls are inflated, and therefore pushed toward the outside (further insights are reported in the Section
on Structural Design).
A rectangular chamber layout is considered, with hand bidentifying the height and breadth of the
chamber section, respectively. In large installations the rectangular layout is generally preferred with
respect to the circular one, because of: i) an easier construction and assembly (the large planar panels
that compose the rectangular section can be easily transported, while circular cylinders are usually
manufactured in a single piece, with increased transport complexity); ii) ease of access and operation
inside the chamber for any technician or experimenter.
Flow-rate Measurement System
The rig configuration selection is also related to the flow-rate measurement system. The rigs in Fig.1
feature in-line orifice-plate systems for the flow-rate measurements. When the flow-rate is lower
or larger than the measurable range the orifice plate must be changed and this operation is time-
consuming. A relatively wide range of measurable flow-rates (0.325 ≤qv≤8m3/s) was required
for the new rig. Thus, a multi-nozzle system appeared as most appropriate, due to the rapidity that
this solution allows in shifting from a nozzle (or a nozzle-combination) to another of different size or
number. A type 3 inlet chamber is indicated in the 2008 version of the ISO Standard [2, p. 114] for
using a multi-nozzle system; thus, the type 3 Inlet Chamber was chosen as final configuration for the
new test rig.
Definition of the Main Dimensions
The design of the rig is an iterative procedure that starts with the definition of the main dimensions.
The facilities for aerodynamic tests are generally quite voluminous: as larger volumes involve higher
costs (space≃money), the test rig should be as small as possible, while permitting tests on larger fans.
The definition of the transversal and longitudinal dimensions of an inlet chamber rig with a multi-
nozzle system (type 3 chamber ) is parametric in terms of:
◦the largest fan inlet diameter (D1max) that is required for the tests;
◦the throat diameter of the largest nozzle (dmax).
Darmstadt (Germany), 18-20 April 2018 4
Sizing of the Transversal Section
The starting point of this design procedure requires the assumption of D1max . Such value normally
comes from the fan manufacturer’s need to test fans up to a specific size. The Standards [1, 2] require
the area of the chamber measurement section A3to be at least 5 times the inlet area A1of the fan under
where A3=h3·b3. Equation 1 allows a preliminary selection of the height and breadth of the chamber
section. In addition, the 2017 Standard [1, pp. 22; 29] imposes two constraints on the dimensions of
b3or h3≥2·D1max (3)
Selecting values of h3and b3that satisfy Eq.s 1, 2, and 3 allows the preliminary sizing of the chamber
section according to the 2017 Standard , which removed the distinction among type 1, 2 and type 3
chambers that was present in previous standards (not only in the 2008 version  but also in the 1998
one). What operationally distinguishes one chamber type from another is unclear and not explained.
Furthermore, additional dimensional constraints were imposed on h3and b3for the type 2 chamber
[2, p. 110], while were not for the type 3 solution. Similar issues exist for the sizing of the rig length,
which is the object of the following sub-section.
Sizing of Longitudinal Dimensions of the Rig
Figure 2 (or the equivalent one from ) is the key to perform the exercise of sizing the rig longitudinal
dimensions. The fore part of the rig (i.e., the part at the left of the multi-nozzle system) is dimensioned
according to Fig. 2 and does not present particular issues, while the minimum longitudinal overall
length lmin of the chamber downstream of the multi-nozzle plane is given by:
lmin ≥largest nozzle length + nozzle-settling means buffer zone +
settling means axial length + settling means-endingwall chamber minimum distance
According to [1, p. 26], all these longitudinal dimensions are parametric in terms of the hydraulic
diameter of the chamber (D3=Dh=4·A
2·(b+h)), whose value derives from the transversal section
sizing, and of the throat diameter of the largest nozzle dmax (see Fig 2). Note that, although not strictly
necessary, it might be advantageous to allow for some space within the chamber to accommodate any
inlet-side equipment (e.g., an inlet-drive motor). Thus, the previous formulation changes into:
lmin ≥largest nozzle length + nozzle-settling means buffer zone +
settling means axial length + settling means-inlet-side fan equipment minimum distance +
inlet-side fan equipment axial length. (4)
For simplicity, the axial length of the inlet-side equipment can be assumed as a multiple of the inlet
diameter of the fan (i.e., =k·D1max , where kis a suitable constant). According to the 2017 Standard
, Eq. 4 translates into:
lmin = 1.6dmax + 2dmax +settling means axial length + 0.5D3+k·D3
Darmstadt (Germany), 18-20 April 2018 5
Figure 2: Axial dimensions of the A-installation type 3 test rig (adapted from Fig. 40e in ).
where the nozzle length (i.e., 1.6dmax) is to be chosen in agreement with the length-to-diameter ratio
[1, pp. 75-76]. The previous version of the Standard  required a larger distance between nozzles
and settling means (2.5dmax, see Fig. 2) than the new Standard  (2dmax), and defines a minimum
distance (= 0.1·Dh) between wire meshes or screens [2, p. 112]. Furthermore, differently from , a
distinction between chamber diameter D3and hydraulic diameter Dhis reported in . Accordingly,
the analogous formulation of Eq. 5 for the 2008 Standard (that was used for the new rig design) is:
lmin = 1.6dmax + 2.5dmax + (0.2Dh+ 3 ·thw)+0.5D3+k·D3
where thwis the wire or screen-plate thickness.
Note that Eq. 6 includes the 2008 Standard distinction between the hydraulic diameter and D3. In
fact, the 2008 Standard [2, p.114] computes the D3diameter with different formulas according to the
type of chamber:
D3=b3·h3type 2 chamber (7)
πtype 3 chamber (8)
It is evident that Eq. 8 provides a D3that is 4
π≃1.13 times the value given by Eq. 7, at equal h3
and b3. Thus, the length of the type 3 inlet chamber is slightly longer than the type 2 chamber.
The measurement section 3 is to be suitably positioned in the chamber, between the settling means and
the first part of the fan inlet-side equipment (see Fig. 2). The knowledge of dmax is required to size the
test rig; however, the design of the multi-nozzle system is a lengthy procedure which has to account
for many requirements (e.g., the accuracy of the pressure sensors, minimum and maximum flow-rates,
continuity of measure within the entire flow-rate range, etc..). It is not reported here because it goes
beyond the scope of this paper. For a preliminary chamber sizing, a reasonable value of dmax can be
chosen (e.g., from similar existing installations), accounting for a limited safety margin.
The transversal and longitudinal sizing of the chamber is thus concluded and the procedure can proceed
to the structural design and verification.
Darmstadt (Germany), 18-20 April 2018 6
Structural Design and Verification
The structural design of the rectangular negative-pressure chamber turned out to be more complex than
the corresponding positive pressure chamber. It was found that there is less interest in rectangular
negative-pressure assemblies in the structural literature than the pressurized case. The information
reported in this Section can mitigate this challenge.
A simple but effective structural layout has been suggested for the new rig: a primary metallic frame
forms the skeleton of the rig, on which the stiffened panels are positioned (see Fig. 3). From the
Figure 3: The metallic frame of the rig that carries the main part of the load and on which the stiffened panels
design perspective, this layout features the advantage of separating the duty required to the different
◦the main loads (e.g., the squeezing action of the pressure) are carried out by the structure along
the beams’ longitudinal direction;
◦the panels (i.e., the chamber walls) transfer the pressure load to the metallic structure. Thus,
they are only required to be stiff enough to maintain the cross section dimensions band hwithin
a given tolerance (see in the following).
While the inlet chamber works almost exclusively under negative pressure loads, some attention must
be provided to the upstream part of the rig (the left side in Fig. 2, ahead of the multi-nozzle system)
that can experience both negative and positive pressure (it is positive when the auxiliary fan is in
The frame must be verified under the maximum static load (e.g., the maximum achievable fan static
pressure, 10000 Pa in our case). In particular, each beam has to be verified against stresses and insta-
◦σmax < ν ·σyi
◦Pmax < ν ·Pcrit
Darmstadt (Germany), 18-20 April 2018 7
Ais the longitudinal stress at the beam section, σyi is the yield stress of the material, ν
is a suitable safety factor and Pcrit is the Eulerian beam instability limit. Suitable references on the
topic can be found in structural handbooks (e.g., ). Particular attention must be provided to the
verification of the junctions among the different beams, as these intersections carry both normal and
Sizing of the Stiffened Panels
The panels must be stiff enough to avoid relevant modifications of the chamber section at the maximum
load condition. The Standard [2, p.26] provides a tolerance on the duct diameter of ±0.01D. Thus, a
quarter of such value (i.e., 0.0025 ·D3) may be taken as preliminary reference for the maximum panel
deformation, as all the chamber sides must be taken into account.
The maximum deflection wmax of a squared simply supported panel subject to an uniformly distributed
load (i.e., a pressure ∆p) occurs at the center, and is given by [9, p.234]
wmax = 0.0443 ·∆pa4
where Eis the Young’s Modulus of the material, ais the panel side length, and thpis the panel
thickness.2At first, in sizing the panel a suitable stiffener position is to be chosen, so defining the side
length a. Secondly, the minimum panel thickness can be computed from Eq. 9 as
The panels that compose the walls of the rig are made of 25 mm poplar plywood, stiffened with 30
mm okumè plywood (see Fig. 4). Further minor verifications of the panels deal with shear stresses at
the bolt positions and at the supporting edges (i.e., where the panel is installed on the metallic frame).
The panel dimensioning concludes the structural design process.
ISSUES RELATED TO THE ISO 5801 STANDARDS
This part of the paper highlights some issues and inconsistencies of the Standards that result in extra
work for the rig designer, involving a deeper analysis of the related flow phenomena and a consequent
interpretation of the Standard.
The main problems encountered during the study that preceded the design of the rig include:
1) the operational difference between type 1-2 and type 3 chambers in the 2008 Standard [2, p. 112]
appears unclear (as a matter of fact, such distinction has been removed from the 2017 Standard).
Beyond the dimensional differences previously mentioned, a second unclear issue is associated
with the total pressure measurement within the type 3 chamber. In particular, it is not clear the
reason why the previous version of the Standard required the mandatory use of the Pitot-tube,
whereas the present version admits also the use of wall pressure tappings.
2The general case for the rectangular plate is reported in . Equation 9 is obtained assuming the Poisson ratio of the
material equal to 0.3
Darmstadt (Germany), 18-20 April 2018 8
2) Both the previous and the new Standard are ambiguous about the positioning of the nozzles with
respect to the chamber axis: they do not clarify if a symmetrical allocation (with respect to the
chamber longitudinal axis) is mandatory (and why) or only recommended.
3) Regarding the surface tolerance of the nozzles, the Standards impose an unclear peak-to-peak
limit on the surface waviness.
For each of the previous items, some insights deriving from the rig design experience are reported and
some improvements are proposed at the end of the Section.
Chamber Type and Use of the Pitot-Tube
The distinction among chamber types was not explained in the previous versions of the ISO 5801
Standard [2, 10]. It is possible that it was still a remnant of the standardization process from the
different national codes and, as a matter of fact, has been removed from the latest Standard version
. Operative differences that were distinguishing the type 3 chamber from the other types were: i)
the allowance of using a multi-nozzle system for flow-rate measurements [2, pp. 114-115], and ii)
the mandatory use of a pitot-tube to measure the air total pressure within the chamber [2, p. 125].
Although the new Standard has removed the imposition of the Pitot tube, allowing the use of wall
static pressure taps within the chamber [1, p. 27], the issue still deserves a brief discussion.
The reasons for imposing the use of the pitot probe to measure the air total pressure pt3within the
chamber were likely to accommodate: i) possible inaccuracies in the area section A3due to the rig
manufacturing and assembly process (that have to be considered when dealing with the typically large
dimensions of such rigs), and ii) possible leakages from the chamber walls that would increase the
flow-rate flowing through the measurement section 3 with respect to the qvmeasured by the multi-
nozzle system. In fact, when wall pressure tappings are used, the total pressure inside the chamber is
where p3is the average static pressure measured at the taps. The computation of the dynamic pressure
term in Eq. 11 is subject to uncertainties on both the flow-rate qvand on the area A3. Leakages from
the chamber walls, when present, shall be identified and minimized (see for instance the Chamber
Leakage Test Procedures in [2, p. 211]). However, uncertainties on A3are quite difficult to be treated,
due to the large dimensions of such assembled structures working under different load conditions.
Such difficulties were likely the reasons to make the pitot solution mandatory in the previous versions
of the Standard. On the other hand, a pitot probe might provide erroneous measurements in presence
of non-uniform velocity profiles within the chamber, caused by ineffective flow-settling means.
The new rig includes both the Pitot-tube sensor and the wall-pressure tappings at the measurement
section, in agreement with . Preliminary comparative tests between the two measurement meth-
ods performed on a 800 mm vane-axial fan running at ∼800 rpm (see Fig.s 6) provided negligible
differences in the measured quantities (i.e., differences in fan pressure pf∼1 Pa).
Symmetrical Positioning of the Nozzles
The ambiguity regarding the positioning of the nozzles for flow-rate measurements is the most rele-
vant, being present in the previous Standards [2, 10] and in the new one  as well.
At p. 69 of  and similarly at p. 75 of the newest code  is reported:
“For tests in standardized airways, multiple nozzles shall be used within inlet or outlet
Darmstadt (Germany), 18-20 April 2018 9
chambers. The nozzles may be of varying sizes but shall be symmetrically positioned
relative to the axis of the chamber, as to both size and radius.”
On the contrary,  at p. 112 reports:
“Multiple nozzles shall be located as symmetrically as possible”
thus allowing for a possible non-symmetrical allocation. The same ambiguity persists in the new
Standard , which shows a non-symmetrical allocation of the nozzles in the figure at page 132.
Apart from such incoherent indications, mostly important is the reason for requiring such symmetrical
nozzle positioning (that is not explained in the norms). The most obvious reason for requiring a sym-
metric allocation is to avoid unbalanced flows within the downstream part of the chamber. However,
this occurrence implies not only the symmetric geometrical allocation of the nozzles (both in size and
radius) but also the simultaneous use of opposite nozzles. This means that equal nozzles at opposite
positions with respect to the chamber axis shall be used simultaneously, to increase the uniformity of
the velocity profile approaching the flow-settling means. However, the authors are aware of fan man-
ufacturers currently using rigs with non-symmetrical nozzle allocation and non-symmetrical operating
(i.e., opened) nozzles.
Evaluating whether the non-symmetrical use of the nozzles might affect the measurements requires a
demanding systematic investigation, which is not currently available to authors’ knowledge. Striving
to ensure a uniform flow at the measurement plane, with the new rig the nozzles:
1. have been positioned symmetrically (to both size and radius) with respect to the axis of the
chamber (see Fig. 4a);
2. are always opened or closed pair by pair (except for the single central one) during a test.
Figure 4: a) The installation of the multi-nozzle wall featuring the symmetrical allocation. b) Metrological
tests to determine whether the nozzle’s surface is in compliance with the tolerances.
Nozzle Surface Waviness Requirement
The third issue on the list deals with a technological aspect of nozzle manufacturing: the allowed
tolerance on the surface waviness for the nozzle’s interior side. As reported previously, the Standards
Darmstadt (Germany), 18-20 April 2018 10
Figure 5: A pressure-decay test performed on the new rig. On the left, the previous smaller inlet-chamber rig
[1, 2, 10] limit the maximum peak-to-peak distance for the surface waviness to one thousandth of the
nozzle throat diameter (i.e., 0.001 ·d). The sense of such requirement is rather mysterious, as finding
a correlation between the peak-to-peak distance and the evolution of the flow-field within the nozzle
is rather difficult. Also from the point of view of technology experts such requirement is problematic,
because it is rather difficult to deal with such waviness tolerances on equipments (i.e., the nozzles)
that are obtained from machining technology.
As nozzle fluid-dynamics is affected by surface smoothness, we hypothesize that such surface tol-
erance is affected by a misprint and the correct requirement would imply a peak-to-valley distance.
Such tolerance would also have an immediate fluid-dynamic explanation, as it would be related to the
typical requirement for smooth duct surface.
The rig’s nozzles were carefully machined, polished and were subjected to different metrological tests,
both dimensional and of surface-characterization (see Fig. 4b)). 3D-laser scans were performed as
well, to determine whether the nozzle profile was in accordance with the nominal one. The results of
the tests confirmed the achievement of the required shape and surface tolerances.
On the basis of the previous discussion, we believe that it would be worth accompanying any dimen-
sional/operative requirement in the Standard with a short explanation. The 2017 code  already
features appreciable steps towards this direction, but there is still margin for improvement.
The rig construction started on May 19, 2017 and was concluded at the end of June 2017. Leakage
tests (see Fig. 5) were performed following both the pressure-decay method [2, p.211] and the two-
phase procedure [2, p.214]. The second procedure (i.e., the two-phase one) turned out to be effective
in identifying the chamber leakages and consequently minimizing them. After the completion of all
the preliminary tests, the rig was put into operation on July 17, testing the 800 mm preswirler-rotor
Darmstadt (Germany), 18-20 April 2018 11
Figure 6: Tests on the 800 mm Preswirler-Rotor fan with the motor on the inlet-side (a) and outlet-side (b).
fan visible in Fig 6. The fan is intended to be driven from the inlet side; accordingly, a 15 kW electric
motor was placed inside the chamber. Seven different tests at different speeds (from ∼400 up to
∼800 rpm) were performed with such inlet-side motor configuration (instrumentation and related
uncertainty are presented in ). As the electric motor is self-cooled by a 80 W axial fan, there were
some worries on eventual detrimental effects on the flow-field inside the chamber and, in turn, on the
fan performance characteristic. As verification, a dedicated outlet-side transmission system has been
manufactured and the tests repeated with the motor outside of the chamber (Fig. 6b)). No appreciable
differences on the fan performance characteristics were observed (i.e., the differences are within the
measurement uncertainty required by the Standard ). While the 800 mm fan size was the maximum
testable on this rig according to , the new constraints imposed by Formulation 3 of the new Standard
 limits the fan size to 700 mm.
Since July 2017 the rig has been used extensively both for research and didactic purposes.
The design and construction of an inlet chamber test rig featuring a multi-nozzle system for flow-rate
measurements has been described. The simple structural layout used for the rig features an internal
frame that supports the stiffened panels of the chamber walls. This layout resulted in an effective and
easy to manufacture assembly, which is now installed at the University of Padova.
The study that was carried out before the rig design highlighted three relevant issues of the ISO 5801
◦the distinction between chamber types and the related (mandatory) use of the pitot probe to
measure the air chamber total pressure;
◦the mandatory/preferred allocation of the nozzles with respect to the axis of the chamber;
◦the tolerance specification on the waviness of the inner surface of the nozzles.
These arguments required time for a deeper investigation during the rig design process; the achieved
Darmstadt (Germany), 18-20 April 2018 12
insights have been reported within the paper. Suggestions for an eventual improvement of a future
Standard have also been provided. In particular, the importance of providing a brief explanation of
the Standard requirements is remarked.
 International Standard Organization. Industrial fans - Performance testing using standardized
airways (ISO 5801:2017),2017.
 International Standard Organization. Industrial fans - Performance testing using standardized
airways (ISO 5801:2008),2008.
 R. A. Wallis. Definition and determination of fan duties. Journal of the Institution of Heating &
Ventilating Engineers, pages 49–59, May 1964.
 AD Martegani, A Lazzaretto, and A Taffurelli. A user’s guide to the UNI 10531 standard on
performance testing of industrial fans. Document ISO/TC117 N, 306, 2003.
 International Standard Organization. Fans - Vocabulary and definitions of categories (ISO
 R Allan Wallis. Axial flow fans and ducts. Krieger, 1993.
 M Masi, S Castegnaro, and A Lazzaretto. Forward sweep to improve the efficiency of rotor-only
tube-axial fans with controlled vortex design blades. Proceedings of the Institution of Mechanical
Engineers, Part A: Journal of Power and Energy, 230(5):512–520, 2016.
 M Spinola, P Gobbato, A Lazzaretto, and M Masi. Effect of Reduced Suction Side Volume on
Cross-Flow Fan Performance. In Proceedings of FAN 2015 Conference. France,2015.
 Thomas Henry Gordon Megson. Aircraft structures for engineering students. Elsevier, 2012.
 International Standard Organization. Industrial fans - Performance testing using standardized
airways (UNI ISO 5801:1998).1998.
 M Masi, A Lazzaretto, and Castegnaro S. Effectiveness of blade forward sweep in a small indus-
trial tube-axial fan. In Proceedings of FAN 2018 Conference, Darmstadt, Germany,2018.
The rig structure has been financed by the University of Padova-Department of Industrial Engineering
with Incentive to Didactics funds. The authors are grateful to prof. M. Carbonaro of the Von Karman
Institute (Belgium) for valuable structural design suggestions, to proff. G. Concheri, R. Meneghello,
and dr. F. Medeossi for performing the surface conditions tests, and to prof. C. Bettanini for his
valuable help with the rig instrumenation. The authors acknowledge Cappellaro Snc and Metalpebo
Srl for their excellent work. The recommendations and suggestions of an anonymous reviewer have
been sincerely appreciated.