The invention concerns a method for tightening screw joints up to
a predetermined pretension level.
In particular, the invention is directed to a process for tightening
screw joints by means of power tool having a torque delivering output shaft and
being connected to a programmable operation control unit by which the rotation
speed of said output shaft is adjustable, and the process includes a primary tightening
phase performed at one or more substantially constant speed levels and a final
tightening phase performed at a variable rotation speed having an average which
is typically lower than the average of the speed levels of the primary tightening
A problem concerned with screw joint tightening is to combine a desireable
short process time with a high accuracy, i.e low scattering as far as the obtained
final pretension level is concerned. Particularly, this is a problem when tightening
varying types of joints, because so called stiff joints, i.e. joints having a steep
torque growth per unit angle of rotation has a tendency to be overtightened due
to a very abrupt arrival at the desired pretension level, whereas so called soft
joints generate a slowly growing torque resistance and arrive at the desired pretension
level after a relatively long retardation period. The basic problem is that the
tightening tool is not able to discontinue its torque application on the joint
quickly enough to avoid overtightening at stiff joints.
An obvious way to reduce scattering and avoid overtightening of stiff
joints is to reduce the rotation speed of the tightening tool such that the target
pretension level is always reached at a lower speed and with a lower kinetic energy
stored in the rotating parts of the tool. Then the tightening tool is able to stop
in time to avoid overtightening. A general reduction of the operation speed of
the tightening tool might have a positive effect on the accuracy and scattering
of the obtained final pretension level. However, a low overall speed is disadvantageous
in that the tightening process would be unacceptably slow, in particular at soft
joints where the torque growth per unit angle of rotation is low.
One previously suggested method for obtaining a quick yet accurate
tightening process, no matter the type of screw joint, is described in US Patent
This prior art method contains a pretensioning step in which the
rotation speed of the tightening tool is successively increased from zero to a
predetermined maximum level. The relatively low speed at the beginning of the
pretension step is intended to prevent overtightening of stiff joints. However,
this succesive speed increase has a predetermined rate and is not adapted to the
actual screw joint characteristic. This means that, even if the speed is successively
increased toward the end of the process, the tightening process as a whole will
be unnecessarily slow for soft joints.
In general, prior art methods utilize torque threshold and/or time
to provide a sufficient speed reduction before the shut-off point in order to control
overshoot caused by inertia and/or occurring sampling intervals.
These strategies have in common that they are based on the assumption
that the joints will have a certain behaviour - Housually a linear torque increase
over time - and that the speed change is based on what is anticipated to happen
later on. Should, however, the joint behave in a non-linear fashion, unexpected
results may occur.
For instance, the method of the above mentioned US Patent No. 5,245,747
is described in connection with three screw joint examples having different torque
rates, and the method is shown to have a compensatory effect on the overshoot
to be expected at the stiffer joints. However, all three joints have linear torque
rates, which means that the method is intended to control overtightening of such
ideal joints where the torque rate characteristic is assumed to be linear all
the way up to the target level. Should instead this method be utilized on a joint
having a non-linear torque rate, for instance a torque rate that is low during
the initial tightening stage and high at the end of the process, the high end speed
would cause a substantial torque overshoot.
The disadvantages of this prior art method is overcome by the method
according to the invention as it is defined in the claims.
Alternative examples of the invention are illustrated in the diagrams
shown on the accompanying drawings wherein:
- Fig. 1 shows a diagram illustrating a single-step tightening process according
to the invention.
- Fig. 2 shows a diagram illustrating a two-step tightening process according
to the invention.
- Fig. 3 shows a diagram illustrating four alternative speed reduction strategies
during the final tightening phase, one of which belongs to prior art.
- Fig. 4 shows a diagram illustrating the different tightening phase durations
for the four alternative speed reduction strategies.
- Fig. 5 shows a diagram illustrating speed adjustment at a part-relaxing screw
- Fig. 6 shows schematically a power tool with an operation control unit for
carrying out the process according to the invention.
The screw joint tightening process illustrated in Fig. 1 comprises
a primary tightening phase I) which is performed at a relatively high speed v1,
and a final tightening phase II) in which the rotation speed is successively reduced
from the primary tightening phase speed v1 to a relatively low final
The primary tightening phase I) includes a running down part and a pretensioning
part, wherein the screw joint is tightened up to a shift torque level TA
which in practical applications is selecetable in the range of 60-80% of the target
torque level TF. Thereafter, the tightening process is completed at
a successively reduced rotation speed to the target torque level TF
where the speed has become v2 as intended.
In this example, the rotation speed v is reduced by a regressively
lower rate, which means that the speed reduction is steepest at the beginning and
becomes successively less steep towards the shut-off point. This speed reduction
is controlled by an algorithm which is using torque as a parameter for executing
the speed control. In particular, this algorithm controls continuously the speed
reduction in response to the relationship between the actual instantaneous transferred
torque T and the target torque level TF.
In this case, the speed at the target torque level TF is
the predetermined variable as is the initial first tightening phase speed v1.
The control algorithm uses torque information obtained from the motor drive system
or from a torque transducer to continuously adapt the rotation speed of the motor
such that the target torque level TF is readched at a predetermined
speed v2 without wasting any time running at this low speed v2
during more than a fraction of the cycle time.
In Fig. 3, there is illustrated four different strategies for reducing
the rotation speed during the final tightening phase II) for arriving at the target
torque level TF at a predetermined low speed v2. The common
aim of all of these strategies is to reduce the speed so as to limit inertial
torque overshoot at the target torque level.
Strategy 1), however, belongs to prior art and means that the speed
is reduced in a single step directly from the first running-down speed level v1
to the intended final speed level v2 as a preselected shift torque level
TA is reached. This means that the entire final tightening phase II)
is performed at a low speed, and a lot of time is wasted.
Strategy 2) means that the speed is reduced progressively steeper
as the actual torque approaches the target torque level, i.e. with a growing steepness.
This means that the speed is upheld as long as possible to reduce the cycle time.
However, this very steep speed reduction at the end of the final tightening phase
II) makes the system very sensitive to torque rate variations between different
joints. There may arrise problems to avoid torque overshoot at very stiff joints,
i.e. joints having a very steep torque growth per angle of rotation.
It is important to keep in mind that the influence of rotation speed
upon inertial contribution to the delivered torque is square related.
Strategy 3) means that the rotation speed is reduced linearly, i.e.
in direct proportion to the difference between the target torque TF
and the actual torque T. Strategy 4) is the one illustrated and described above
in connection with the process illustrated in Fig.1 and means that the rotation
speed is low during a longer interval before reaching the target torque level as
compared to the linear strategy 3). This means just a little slower process but
also that the system is less sensitive to torque rate variations between different
joints. Put together, this means that a safer overshoot control is obtained substantially
without causing any longer cycle time.
As illustrated in Fig.4, the time intervals spent by the four alternative
speed reduction strategies to reach the target torque level differs a lot. The
previously known constant speed alternative 1) takes the longest time and causes
an undesireable long cycle time. The process is not completed until t1.
Strategy 2) on the other hand, results in a very fast process ending
at t2 where the target torque is reached.
Strategies 3) and 4) are both faster than strategy 1) but slower
than strategy 2). Strategy 4) in particular, combines a safe tightening process
as regards the risk for inertial overshoot with a reasonable short final tightening
From the ergonomic point of view, speed reduction in accordance with
the invention results in a lower reaction force exposure on the operator, since
the human muscular control system prefers smooth load changes to abrupt ones.
This means that the speed reduction strategies comprised in the concept of this
invention also give a favourable reaction torque characteristic to be handled by
the operator. From this point of view, it is apparent that the prior art strategy
1) is less favourable in that it contains a knee in the torque growth characteristic
dM/dt caused by the sudden speed change at the shift torque level TA.
This causes a discontinuity in the reaction force which is awkward and tiresome
for the operator to cope with.
The tightening process illustrated in the diagram in Fig.2 comprises
a primary tightening phase I) which is divided into two steps IA and
IB performed at different rotation speed levels. Accordingly, the first
step IA is performed at a relatively high speed v1A and ends
as the delivered torque reaches a snug level TS. From that point on
the second step IB takes over and continues the primary tightening
phase at the lower speed v1B.
By dividing the primary tightening phase into two steps with different
rotation speed there is obtained a shorter cycle time due to a high speed initial
running down of the joint without increasing the risk for torque overshoot at
very stiff joints.
To construct algorithms providing successive speed reductions of
different characteristics within the scope of the invention and as described above
is a straightforward mathematical task and can be done in several different ways.
For instance, the speed v can be determined by comparing the product
of the intended target speed v2 and the quotient of the target torque
TF and the actual instantaneous torque T with the v2 value.
As the actual torque T increases, the value of TF / T x v2
approaches the value of v2, i.e. the quotient TF / T gets
closer to 1 and the speed is aimed at v2.
The initial constant speed v1 in the process illustrated
in Fig.1 is determined by the maximum speed of the motor drive system and is not
a result of the speed controlling algorithm. However, as the speed control value
of the algorithm becomes lower than v1, the speed reduction is started.
The tightening process according to the invention differs from prior
art methods in a distinct way in that the rotation speed during the tightening
phase is controlled in response to the actual difference between the actual instantaneous
torque level T and the target torque level TF. This means that should
the torque resistance in the screw joint decrease during a certain interval, the
rotation speed would be increased momentarily during that interval, because during
that interval the difference between the actual instantaneous torque level T and
the target torque level TF would increase. At the end, the torque resistance
will always increase to reach the target torque level TF. Such a process
is illustrated in Fig.5.
This successive adaptation of the rotation speed in response to the
actual span between the actual instantaneous torque level T and the target torque
gives the advantage of keeping up the tightening speed without
risking an approach to the final torque level TF at too a high speed.
In Fig 6, there is shown by way of example an equipment for carrying
out the invention. This equipment comprises a portable electric power nutrunning
tool 10 having a rotating output shaft 11 and a handle 12. The tool 10 is connected
to an operation control unit CU via a multi-core cable 13. The control unit CU
includes means for supplying electric power to the tool 10 and a programmable means
for govering the power supply to the tool 10 in respect of parameters like voltage,
current and frequency.
An example of a suitable control unit is a system commercially available under
the name: POWER FOCUS 2000 marketed by Atlas Copco.