Design of a compact six-component force and moment

sensor for aerodynamic testing

Marin SANDU*, Adriana SANDU*, Georgeta IONACU*

*Corresponding author
POLITEHNICA University of Bucharest,
Splaiul Independenei 313, 060042, Bucharest, Romania
marin_sandu@yahoo.com
DOI: 10.13111/2066-8201.2011.3.1.13

Abstract: The measurement of steady and fluctuating forces acting on a body in a flow is one of the
main tasks in wind-tunnel experiments. Usually, a multi-component strain gauge force and moment
sensor (also known as balance) is used to generate signals which are processed by means of an
adequate instrumentation.
To design a wind-tunnel balance, the specifications of the load ranges and the available space (for the
placement of the balance inside or outside the model) are required. The main challenge is to conceive
the elastic element of the sensor as a monolithic part with a relative simple geometry and to identify
the adequate placement of strain gauges to maximize the measuring sensitivities and to diminish the
inter-influence of the components.
This paper describes the design of a six-component force/moment sensor which is compact, has high
measuring sensitivities, and can be used either as internal or as external balance in the aerodynamic
testing.
Key Words: six-component sensor, wind-tunnel balance, design procedure

1. INTRODUCTION
The force and moment strain gauge transducers have important applications in
Aeromechanics, Robotics, Automatic Control and Monitoring Systems, and also in
Biomechanics. Since the force and moment are vectorial physical quantities, their
magnitudes and directions are used to define them. As is well known, a mechanical load can
be divided into six components: three force components and three moment components.
Consequently, to define completely the load, it is necessary to know the values of all six
components. A multi-component force/moment sensor is a device that enables forces and
moments to be measured simultaneously. If they are placed inside of the model they are
referred to as internal balances and if they are located outside of the model or the wind
tunnel, they are referred to as external balances. There is limited space inside of the model
itself, so internal balances have to be relatively small in comparison to external balances.
There are two main types of internal balances. The monolithic type, in which the
balance body consists of a single piece, properly designed such that certain areas are
primarily stressed by the applied loads. The other internal balance type uses six small
transducers which are oriented with their sensing axes in the direction of the applied loads.
Such a balance is combined into a solid structure. A balance measures the total load acting
on the model and therefore is placed at the center of gravity of the model.
The six different components of aerodynamic loads (three forces in the directions of the
coordinate axes and the moments around these axes themselves), are measured in a certain
coordinate system which can be either fixed to the model or to the wind tunnel.
INCAS BULLETIN, Volume 3, Issue 1/ 2011, pp. 95 100 ISSN 2066 8201
 Marin SANDU, Adriana SANDU, Georgeta IONACU 96

The basic metrological features of multi-component strain gauge sensors are determined
to a great extent by the shapes and sizes of elastic elements on which the strain gauges are
bonded [1]-[8].
The geometry of the elastic element of the sensor must comply with several
metrological requirements concerning parameters as: nominal loads, allowed overload,
measuring sensitivities, rigidities, linearity, hysteretic effects, interaction between the
measured components, eigenfrequencies, fatigue life.

2. DESIGN DATA AND GEOMETRY OF THE ELASTIC ELEMENT

The components which act on the model and are required to be measured are the following:
axial force Fxm , side force Fym , normal force Fzm , rolling moment M xm , pitching moment
M ym , yawing moment M zm . According to the standard ISO 1151 their positive directions
are considered as in figure 1. A balance which stays fixed in the tunnel, being related to the
wind-tunnel axis system, always gives the pure aerodynamic loads on the model. In the case
of the axis system attached on model, the balance does not measure the aerodynamic loads
directly. The loads acting on the model are given by the balance and the pure aerodynamic
loads must then be calculated from these components using the correct yaw and pitch angles.

Fig. 1 Definition of model-fixed axis system according to standard ISO 1151

To choose the configuration of an elastic element, some general recommendations are

taken in consideration such as:
symmetric structures are preferred because it is possible to eliminate the cross influences
of the load components by an adequate positioning of strain gauges,
the signals can be amplified connecting in Wheatstone bridges pairs of stretched and
compressed strain gauges,
the elastic element will be designed as a monolithic structure in order to reduce the
nonlinearities and the hysteretic effects which can be induced if assembled parts with
clearance are used.
The elastic element of the sensor that will be discussed in this paper is a relative simple
beam structure (Fig. 2), made from an alloyed steel (34MoCrNi15) that has the following
characteristics: ultimate strength Rm 1300 MPa, yield limit R p 0, 2 1000 MPa, Youngs
modulus E 2 105 MPa and Poissons ratio 0.3 .

INCAS BULLETIN, Volume 3, Issue 1/ 2011

97 Design of a compact six-component force and moment sensor for aerodynamic testing

The values of nominal loads which will be considered in the origin of the global system
OXYZ are: Fxm = Fym = 500 N, Fzm = 1000 N, M xm = M ym = M zm 510 4 Nmm.
Other design requirements were the following:
- external sizes of the sensor (mm): 80 120,
- measuring sensitivities greater than 0.5 mV/V (for a supply voltage of 10 V),
- cross influences between the measured components: max. 8%.
In order to design an elastic element with convenient external sizes, the lengths of the
beams were taken as: l1 50 mm, l2 35 mm, l3 20 mm.

Fig. 2 The calculus model of the sensor elastic structure

All the beams have the same square cross section a a . Using the well known method
of the forces [7], an analytic pre-dimensioning calculus, that will not be presented here, was
performed for the case when the components Fxm , Fzm and M ym act simultaneously and the
allowed maximum stress in the structure is a R p 0, 2 / 2 =500 MPa. As a result it was
established that the convenient dimension is a 8 mm. The combined loading case that was
taken into account in the pre-dimensioning stage is corresponding, at the model level, to the
aircraft take-off.

3. FINITE ELEMENT ANALYSES, PLACEMENT OF STRAIN GAUGES

AND EVALUATION OF MEASURING SENSITIVITIES
Six cases, corresponding to successive loading with the nominal values of forces and
moments were considered, i.e the load applied in the origin of global reference system OXYZ
(Fig. 2) was: 1) Fxm , 2) Fym , 3) Fzm , 4) M xm , 5) M ym , 6) M zm . The hypothetic case 7 when all
the loads which are applied in cases 1 to 6 act simultaneously will be obtained by
superposing the effects. The point O is placed on the middle of a fictive and more rigid beam
that materialize the joint between the model and the sensor. At the opposite end, the elastic
structure is fixed in the tail sting which sustains the aircraft model in the wind tunnel.

INCAS BULLETIN, Volume 3, Issue 1/ 2011

 Marin SANDU, Adriana SANDU, Georgeta IONACU 98

The main results of finite element analyses are the bending moment diagrams (fig. 3) which
are reported to the local reference systems proposed for each component beam (Fig. 2).
Because the beams have a massive section, the effect of bending moments is dominant and
the contributions of axial and shear forces can be neglected.

Load case 1 Load case 2

Load case 3 Load case 4

Load case 5 Load case 6

Fig. 3 The bending moments diagrams (values in Nmm)

INCAS BULLETIN, Volume 3, Issue 1/ 2011

 99 Design of a compact six-component force and moment sensor for aerodynamic testing

Also, because the strain gauges will be bonded in longitudinal direction on the beams,
their indications will be not influenced by the torques. However, the effect of the torques was
taken into account in the evaluation of maximum equivalent stress ( eq , max ) in the structure.
The strength condition ( eq , max a ) was accomplished in all loading cases 1 to 6.
Consequently, the dimensions of elastic element are convenient, although in the practically
improbable case 7, the maximum equivalent stress is of about 800 MPa.
Taking into account the symmetry/antisymmetry of the diagrams from figure 3 and the
presence of some cross sections with zero or very little values of bending moments, the
adequate positioning of strain gauges was identified as in figure 4.

Fig. 4 Placement of strain gauges on the elastic element and the measuring Wheatstone bridges

The shape of elastic elements for multi-component sensors is conceived in order to

ensure the independence of the signals which are proportional /to the measured components.
In the ideal situation when the influence between effects is totally eliminated, a significant
signal is obtained for the measured component while the signals for the other components
are zero.
When the effect are coupled, a coupling matrix [Cij ] is experimentally established by
calibrating the sensor. For a six component sensor, the following relation can be written
{Si } [Cij ]{L j } , (i, j 1, 2, ... ,6 ) , (1)
where [Cij ] is a 66 matrix, {Si } is the output signal vector for the six measuring bridges
and {L j } is the load vector ( L1 Fx , L2 Fy , L3 Fz , L4 M x , L5 M y , L6 M z ).
The relation obtained from (1) after multiplication to the left by [Cij ]1
{L j } [Cij ]1 {Si } , (i, j 1, 2, ... ,6 ) , (2)

INCAS BULLETIN, Volume 3, Issue 1/ 2011

 Marin SANDU, Adriana SANDU, Georgeta IONACU 100

will be used to determine the unknown loads that act on the model, based on the signals
recorded during its testing in the wind-tunnel.
In the particular case of total independence of the six signals, [Cij ] is a diagonal matrix.
Usually, because of manufacturing imperfections and of strain gauges positioning errors, the
matrix will have most of its elements different from zero, but the elements of the main
diagonal will be larger than the others.
The signal provided by a full Wheatstone bridge (Fig. 4), corresponding to any
component can be evaluated by the relation
U e (kt / 4) (1 2 3 4 ) U a , (3)
where kt is the constant of strain gauge (usually kt =2), i (i=1, 2, 3, 4) are the strains
captured by the strain gauges T1 , T2 , T3 , T4 and U a is the measuring bridge supply voltage.
The measuring sensitivity of each Wheatstone bridge can be determined by the formula
S m U e / U a . (4)
Based on the bending moments diagrams (Fig. 3) and taking into account the strain
gauges positions (Fig. 4), measuring sensitivities ranging from 0.7 mV/V to 2.9 mV/V were
estimated for the six components. The maximum coupling error was identified between the
components Fy and Mz , being of about 5%.

4. CONCLUSIONS
The six-axes force/moment sensor that was analysed in this paper has a compact
configuration and it can be used either as internal or as external balance in the aerodynamic
testing of aircraft models in the wind-tunnel. Also, since the shape of elastic element is
simple, an accurate manufacturing with a low cost price is possible. The estimated
metrological characteristics of the sensor are convenient for usual measurements.

REFERENCES
[1] I. Constantinescu, D. M. tefnescu, M. Sandu, Msurarea mrimilor mecanice cu ajutorul tensometriei,
Bucureti, Editura Tehnic, 1989
[2] J. W. Dally, W. F. Riley, K. G. McConnell, Instrumentation for engineering measurements, New York, USA,
John Wiley & Sons, 2nd ed., 1984
[3] G. S. Kim, Development of a six-axis force/moment sensor with rectangular taper beams for an intelligent
robot, International Journal of Control, Automation, and Systems, vol. 5, no. 4, pp. 419-428, 2007
[4] A. Marinescu, Metode, aparate i instalaii de msur n aeromecanic, Editura Academiei Romne,
Bucureti, 1970
[5] Y. K. Rark, R. Kumme, D. Roeske, D. I. Kang, Column-type multi-component force transducers and their
evaluation for dynamic measurement, Measurement Science and Technology, vol. 19, pp. 1-10, 2008
[6] A. Sandu, M. Sandu, C. Atanasiu, A low-profile sensor for six force/torque components, pp. 67-69,
Proceedings of the 18th DANUBIA-ADRIA Symposium on Experimental Methods in Solid Mechanics,
Steyr, Austria, September 26-29, 2001
[7] M. Sandu, A. Sandu, Analytic-numerical approach in a multicomponent strain gauge transducer design,
Scientific Bulletin of University POLITEHNICA of Bucharest, vol 67, no. 3, pp. 37- 44, 2005
[8] M. Sandu, A. Sandu, M. Gvan, Consideraii privind proiectarea balanelor tensometrice de suflerie,
Aerospace International Symposium Carafoli 2001, pag. 169-174, Bucharest, 2001
[9] * * * COSMOS/M Finite Element System, User Guide, Structural Research & Analysis Co.,USA, 1998

INCAS BULLETIN, Volume 3, Issue 1/ 2011