19th World Conference on No

License: http://creativecommons.org/license

Tighte

Farid1 UL

Abstract. Control of different sectors, wheheterogeneous tighteni To avoid or mitigcontrol of tightening sealing as well as assedynamic solicitations. This paper descprinciple of the ultraso This technology ultrasonic wave alongsensor is thus positionelive. The information othe screw allows meas

1. Introduction

One of the major problems achieving an accurate preloadthe detachment of the assemany additional stresses due to

Insufficient, excessive or hetand bad tightening conditionsproblem in all industries: trastrong financial and human process with precise and reliaThere was a strong need to ewith the aim to better contrassemblies. Portable or transpto these needs, and are nowstress, elongation, and angle) To control the tightening of ththe use of torque wrench anscrew, rods) using either meultrasonic method. The torqueshow enormous intrinsic error

Non-Destructive Testing 2016

1 nses/by/3.0/

htening control by ultrasound

arid BELAHCENE 1, Pierre SAMSON 1 ULTRA RS, 10450 Breviandes, France

Contact e-mail: f.belahcene@ultrars.com

of screws and bolts tightening takes a growing imporhere assembly defects caused by insufficient, exce

ening could induce hazardous consequences. itigate the failures related to tightening, the measuremeng pretension should be performed in order to ensuressembly stiffness which allow withstanding external sts.

escribes a new ultrasonic method of the tightening asonic measurement and some experimental results. gy is based on the measurement of the travel timeng the elongated screw during the tightening. An uloned inside a socket in order to measure the screw traven of the travel time combined with the calibration coeffasuring the actual tightening force applied.

with the use of bolted joints is the precisiooad. It is essential to apply an adapted tighteninembled parts under the influence of external foto vibration, shock...

heterogeneous tightening represent about 30%ons represent about 45% failures in fatigue. Thtransport, oil and gas, nuclear, shipbuilding

an impact. Therefore, it is necessary to contrliable measurement tools. o evaluate the tightening force in bolted jointsntrol the assembly processes and increase thnsportable instruments have been developed toow commercially available. They provide info

) on the quality of the assembly. f the assembled parts, there are different metho and methods that determine the elongation omechanical strain gauges or more sophisticateque wrench method adapted for this purpose harrors in bolt tension measurement by up to 30%

portance in cessive or

ent and the ure perfect l static and

ng control,

ime of the ultrasound vel time in efficient of

sion, with regard to ing force to prevent l forces, and nullify

% of static failures They are a recurrent g etc ..., causing a ntrol the tightening

ts non-destructively the lifetime of the to provide solutions

information (torque,

thods. These include n of fasteners (bolt, ated means such as has been reported to %. It’s dissipated in

More info about this article: http://ndt.net/?id=19681

2

overcoming friction under the bolt head or the nut face (whichever face that is rotated). Identical bolts, when tightened to identical torque values, vary substantially in their actual tensions [1]. The ultrasonic method is independent on the friction. This technique is accurate, non-destructive, easy to implement, and can be used to control the tightening in real time. However, the limits of the ultrasonic method (single wavelength) are imposed by the need of the original length of the bolts before tightening and material constants such as Young's modulus to determine the actual tensile load from 'ultrasonic elongation' of the bolts.

2. Principal of the ultrasonic stress measurement

The linear theory of elasticity is generally adequate to describe the elastic behaviour of materials, by using Hooke’s law. In this approach, the elastic strain energy is developed to the second-order terms for isotropic media. However, the theoretical description of the acoustoelastic effect, which refers to the change in velocity of ultrasonic waves propagating in strained solids, is only possible by considering the non-linear theory of elasticity. Using Murnaghan theory [2], which takes into account third-order terms in the strain energy, Hughes and Kelly derived expressions for the velocities of elastic waves in a stressed solid [3]; in the following form:

2312

130

3212

120

12

110

21

24)(

21

24)(

)1044()2(2

εµεµεθλµρ

εµεµεθλµρ

εµλθλµλρ

nmV

nmV

mlV

−++++=

−++++=

++++++=

(1)

Where:

• V1j is the velocity of waves in the direction 1 with particle displacement in the j direction,

• ρ0 is the initial density, • 321 ε+ε+ε=θ are the components of the homogeneous triaxial principal strains, • λ etµ : second order elastic constants, • n,m,l : Third order elastic constants.

In the case of uniaxial stress, ε ε1 = andε ε νε2 3= = − , where ν is the Poisson ratio. By linearizing the equation (1) in the first order, we can write: ( )111011 1 σjjj AVV += (2) Where:

• V1j°: velocity of the wave in the direction 1 with zero stress, • σ11: stress in the direction 1, • V1j: velocity of the wave in the direction 1, in the presence of stresses

(σ11), • A1j: acoustoelastic constant.

Application of the acoustoela

Figure 1 represents the configorder to, an ultrasonic signalthrough the screw until reacreturns to its head. The time of the screw.

Fig. 1. Co

When a screw is tightened, twFirstly, the acoustoelastic effeIt tends to decrease the propatime. Secondly, the effect ofincreases the travel time.

We consider a simple one-diaxial load (Figure 2). The totwave is V0 in the stress-free head. When the tensile stress σ is apelongated to (1+σ/E).dx, Assuming that the wave veloc

V =

V0 is the wave velocity witho

The travel time in the small se

dt =

The higher-order terms are ne

The round trip travel time alfollowing formula:

3

lastic theory for the tightening measurement

figuration of the tightening tension measuremennal is applied on the head of the screw. The aching the air / steel interface at the end of e between the 1st and the 2nd echo is used to m

Configuration of the tightening tension measurement.

, two effects will increase the travel time of theeffect, which is caused directly by the tensile stopagation velocity of the wave and, therefore, of lengthening of the screw increases the wa

dimensional model to calculate the plane-wav total length of the bolt without stress is L0 andee conditions; lN and lH are the lengths of the

applied along the length, a small segment dx in, where E is the Young modulus olocity V varies linearly with axial stress σ, that

).1(0 σKV += (3) hout stress, and K is the acoustoelastic constant

l segment dx becomes:

( )[ ] dxVKE .1 101 −− −+= σ (4) neglected in the derivation.

along the bolt is obtained by integration, and

ent:

ent in the screw. In e signal propagates of its rod, and then measure the length

the ultrasonic wave. stress in the screw. e, increase its travel wave path and thus

ave response to the and the longitudinal the nut and the bolt

in the bolt shank is of the material.

at is:

ant.

and is given by the

t = The formula (5) can also be w

t

t

0

−

Here:

0

00

2

V

Lt =

ECB =

−1

CB is the calibration coeffielongation effect of the screw The effective length ratio β is

=β

Fig. 2. One-dimentional

3. Influences of external par

3.1 The effect of the variation

The calibration coefficient dea non-negligible influence onaccuracy of the measurement Figure 3 represent an examp(M42) with two different proportionality between the rapplied stress (CB) ranges respectively for Ls of 54 mm

4

).1(0 aveBCt σ+ (5)

e written as follows:

aveBCt σ.

0

0 =−

(6)

K−

max.σβσ =ave

fficient which depends on the acoustoelastiew.

is defined as:

0.21

L

ll HN +−=

(7)

al axial-stress model [4].

parameters on the ultrasonic measurement

ion of the tightened length

depends on the tight length (Ls). The change of on the results of calibration coefficient CB andnt of the tightening tension.

mple of the calibration coefficient CB obtainedt values of Ls. This result shows that thee relative change in travel time of the longitudis from 1.48 10-5 to 1.2 10-5 MPa-1 (about m and 40 mm.

stic effect and the

of Ls (Figure 2) has and therefore on the

ed on a steel screw the relationship of udinal wave and the ut 20% difference)

The formula (6) requires thecoefficient CB and an accurate

Fig. 3. Calibration coeffici

3.2 The effect of the variation

The temperature is an impomeasuring the tightening strvelocities of the ultrasonic waThe velocity of the ultrasonic

ρ=V

Where V is the velocity, E is It is clear that E and ρ are depinfluences the velocity of thmeasured. A linear increase in temperaexpressed in MPa). For exam8 MPa per degree in the rod MTo correct the temperature eff

Where:

avemes σσ = (formula 6)

T°: initial temperature during

T: temperature at time t

Therefore, the measurement otravel time of ultrasonic wavdepending on the temperature

5

the knowledge of the right value of Ls to orate tightening stress.

ficient CB obtained for two tight lengths Ls [1].

ion of the temperature

portant factor on the application of the ultrasstress. The influence of the temperature on

waves is well known [5]. nic wave, (independently of the tightening stress

)21)(1(

)1(

ννρν−+

−E

(8)

is Young's modulus, ν is the Poisson's ratio andependent of the temperature. Therefore, a chanf the wave and changes the accuracy of the

erature causes a linear increase of the measuample, the stress measured by the longitudinal wd M48*300 - 34CrNiMo6. effect on the measurement, the stress equation

)(8 TTmesreel −°+= σσ

(9)

ng the first measurement

nt of the tightening stress on the screw is realizaves propagating on the screw and correctin

ure of the component under evaluation.

obtain an accurate

trasonic method for on the propagation

ess) is defined as:

and ρ is the density. ange in temperature he tightening stress

asured stress (error al wave increases by

can be rewritten:

lized by reading the ting the travel time

6

3.3 The effect of the parasitic bending

The bending stress can be caused by a variation of the thickness of a washer or the presence of a weld point under the head of the screw. The ultrasonic measurement of the tightening stress will be erroneous if the screw bends: the ultrasonic signal will be influenced by the presence of the bend in the screw. Table 1 represent an example of measurement obtained on a steel screw (M16x60) with a washer of inconstant thickness. A variation in the thickness of a washer of 1 mm may induce a bend in the screw and result in overestimation of stress measured by ultrasound. For this example, the relative difference between strain gauge and ultrasound is equal to 11 %.

Torque applied with a torque wrench (N.m)

Stress measured by strain gauge (MPa)

Stress measured by ultrasound (MPa)

With bending stress

0 0 0 130 40,8 45,46 11,4

Without bending stress 130 N.m 36,4 35,75 -1,8 130 N.m 318 31,59 2,6 130 N.m 32,9 33,3 1,2

Table 1. Stress measured in a steel screw M16 x 60 - effect of the variation of the thickness of a washer.

4. Calibration Procedure

The calibration is necessary to know the coefficient CB connecting the measured travel time of the ultrasonic waves to the imposed tightening stress. This coefficient depends on the screw diameter, the tight length Ls and mechanical characteristics of the assembly. For a given tightening configuration (diameter, tensile length Ls, materials), we must perform a calibration to determine the coefficient CB. The screw is tightened in the steel block (figure 4) and increasing levels of different loads are applied (step by step) with an hydraulic cylinder and measured with a force transducer. The ultrasonic sensor is placed on the screw head to measure the relative change in travel time of the longitudinal wave. An example of calibration is represented above in figure 3.

SG

USSG

F

FFerror

−=(%)

5. Tightening Stress Procedu

The ultrasonic device used tconsists of:

- the hydraulic torque - the socket instrumen- and the ultrasonic de

In order to optimize the tightthe technique of hydraulic tiand rapid homogeneous tighteTo address this problem, a pautomated system: hydraulic as a means of stopping the hythe tightening force in the tightening the screw until the In addition, the tightening opeThis provides a uniform tighte Figure 5 represent the mock-The calibration coefficients oThe screw is tightened to a strain gauge (which is conside

Screw

Hydraulic cylinder

7

Fig. 4. Experimental mock up for the calibration.

edure

d to control the tightening stress is represente

ue wrench piloted by ultrasonic device, ented by ultrasonic sensor, device.

htening operation, the coupling of the ultrasontightening would be interesting in order to en

htening. a project was born, which involved the develolic torque wrench piloted by ultrasonic devicehydraulic tightening. The operator will be able software; the hydraulic torque wrench will

he desired value is obtained.

operations can be performed simultaneously onhtening force on all screws concerned (homoge

up used to qualify this technique. s of the screws were previously determined. a specified load and the actual force is measidered as the reference) and the US equipment (

Ftr

Ultrso

nted in figure 5. It

onic technique with ensure an accurate

elopment of a fully ce: using ultrasound ble to fix a value of ill then proceed to

on a several screws. genous assembly).

easured by both the nt (FUS).

Force transducer

Washer

Steel block

ltrasonic socket

Fig. 5. Experimen

As shown in the fig.6 and tabspecified force.

Target value

50

100

150

200

250

300

50

100

150

200

250

300

50

100

150

200

250

300

The tightening procedure is: software, the hydraulic torquthe tightening when the ultras

8

ental setup for measurement of tightening force.

table 2 this method enables to accurately tighte

FSG (kN) FUS (kN)

50,3 49,0 2,6%

100,7 99,5 1,2%

150,5 149,1 0,9%

200,3 200,5 0,1%

247,5 249,9 1,0%

295,2 299,6 1,5%

51,6 49,3 4,5%

100,2 99,8 0,4%

152 150,6 0,9%

200 199,9 0,0%

248,2 250,4 0,9%

298,2 300,0 0,6%

50,7 49,0 3,4%

103 100,1 2,8%

153,5 151,2 1,5%

203,7 201,3 1,2%

253,7 252,4 0,5%

301,3 301,5 0,1% Table 2. Stress measured steel screw M42 x 300.

is: the operator fix a target value of the tighterque wrench will then proceed to tightening thrasonic force value (FUS) is reach the target valu

SG

F

Ferror =(%)

hten a screw up to a

htening force in the the screw and stop lue.

SG

US

F

F−

Fig. 6. Tightenin As show in the table 2, the vtarget values.

6. Conclusion

The accurate control of the tdesign of mechanical equiptightening is the cause of abothe cause about 45% of failur

Therefore, the tightening opprecise, reliable, easy to impwith a minimum of operation

Ultrasound is one of the techntightening force is providedmethodology for measuring th

The choices that were made a- The use of specific ultrasoni- Calibrations of acoustoelasti-The use of the UltraRS soforder to minimize the error an

References

[1] BELAHCENE (F). - Contrôle dR4040 | Date de publication :10 nov[2] MURNAGHAN (D.). - Finite d[3] HUGUES (D.S.) et KELLY (J.no 5,p. 1145 à 1149 (1953). [4] Masahiko (H) and Hirotsugmeasurements. Kluwer Academic P[5] ZHOU (X). – Etude paramétrultrasonore,Thèse de doctorat, UTT

9

ning at ambient temperature - Error versus tightening for

values of the forces measured by ultrasound

e tightening force is often seen as a secondaryipment, and yet the insufficient, excessive bout 30% of static failures and the bad tightenilures by fatigue.

operation must be controlled through measuimplement, enabling a homogeneous tighteninons.

chniques to address this issue: the ultrasonic deted by experimental devices that were designg the travel time (speed) of the ultrasonic wave.

e at the experimental elements are: onic transducer to obtain better reproducibility astic effect for selecting the longitudinal wave fsoftware for the effective treatment of measur and improve the acquisition speed.

le de tension de serrage par ultrasons - Techniques de lnov. 2015. e deformation of an elastic solid, John Willey, New York (J.L.). - Second-order elastic deformation of solid, Phys

ugu (O). -EMATS for science and industry: noncic Publishers, Boston/Dordrecht/London 2003. étriques pour la détermination des contraintes résidueTT, France (2006).

force.

nd are very close to

ary operation in the e or heterogeneous ening conditions are

asuring instruments, ing of all the bolts

determination of the igned to refine the

ve.

of measurements, for measurement,

surement signals in

e l'ingénieur, Référence

ork (1951). hysical Review, vol. 92,

oncontacting ultrasonic

duelles par la méthode