Measurement 114 (2018) 208–217

Contents lists available at ScienceDirect

Measurement
journal homepage: www.elsevier.com/locate/measurement

A case study in technological quality assurance of a metric screw thread MARK

A R T I C L E I N F O A B S T R A C T

Keywords: This paper presents a set of actions taken within the quality system in order to ensure that a product fulfils
Cyber physical system quality requirements. The quality of the product was guaranteed by application of the concept of cyber physical
Metric screw thread system using the three quality management stages strategy proposed by Taguchi. Technological quality assur-
Quality assurance ance of a product was proposed on the example on a metric screw thread. The paper is divided into three parts.
Dimensional and form accuracy
Firstly, the specifications of a product, technology and measurement system are introduced, then the capability
Blade wear
MSA analysis
of the system is tested, and finally the controllable and uncontrollable factors influence on the system are
contemplated with regard to providing the product with technological quality.

1. Introduction

Quality assurance is defined as activities implemented within the quality system that can be demonstrated to ensure that a product will fulfil
requirements with regard to quality. Quality itself can be related to reduction of variation, value-added, and conformance to specifications.
Making a product in accordance with the assumed tolerance requires the application of the correct manufacturing process. One can distinguish
the manufacturing process and the production system constituting the environment for the manufacturing process. The accuracy of the process can
be affected by many elements, such as: the workpiece, instruments and machines, machining method, production operator and organisation. All
these elements, as well as other unmentioned ones, influence the real value range of characteristics of the process. In consequence, the produced
quality is different from the assumed. The variability of the produced quality (precision) denotes repeatability and reproducibility of the product's
manufacturing.
According to the tests conducted by [4], quality providing in manufacturing takes place in production facilities only to a limited extent, mainly
due to the complicated nature of the procedures involving capability assessment of the measurement system. Therefore, in [4] a simplified procedure
involving the implementation of ongoing assessment of the measurement system in manufacturing was proposed, in which operators perform
measurements as part of their daily work at their work stations, without marking the samples. The proposed procedure of ongoing assessment of the
measurement system means minimum preparatory work input, without intervening into the manufacturing process and operators' work. As far as the
product's manufacturing process is concerned, the basic matter is providing technological quality defined by the values of dimensions, contours of
circularity, cylindricity and shape, as well as roughness and waviness [1,8,19]. This set of properties, described with parameters, defines the
properties of the product and its suitability.
In order to be able to provide the machined product with the technological quality assumed in specifications, it is necessary for the production
system to have the appropriate quality providing capability in a given tolerance, and a measurement system able to verify it (Fig. 1a). After the
process design the requirement towards quality assurance are sent to manufacturing system, which provide all the necessary elements to produce the
metric screw thread within required tolerance.
Nowadays it is not sufficient to provide the measurement process to ensure the quality of the product. It is necessary to apply metrology for
worry-free productivity created within the cyber-physical system (Fig. 1b). After the process design the supervising of the quality of metric screw
thread is realised in cyber-physical system. The quality requirement relies upon a mathematical model describing the relations between metrological
and technological parameters. The technological parameters are sent to manufacturing system, where data from the sensors (metrological para-
meters) are obtained and analysed for mathematical model to be improved. In this way the convergence between the required and achieved
parameters is possible to attain and the quality of metric screw thread is ensured.
In order to obtain a product characterised by given properties, it is necessary to define the limitations concerning those properties. In the case of
geometrical limitations, they are tolerances for every defined surface of the product. The tolerances are taken into consideration during preparations
for the production; at that time, design quality and technology of the product are defined, and the conditions for providing the appropriate product
quality and technology implementation are created.
During production development, where product quality is controlled off-line, it is possible to adopt the strategy proposed by Taguchi [10]. It
consists of the following three quality management stages in designing:

- Stage 1, system design, consists in preparing product specifications and the concept of the manufacturing system.
- Stage 2, parameter design, consists in performing preliminary tests on nominal values and, on that basis, determining the closest combinations of
parameter levels and operative levels of the process indices, so that they are most resistant to changes in the environment and other confounding

http://dx.doi.org/10.1016/j.measurement.2017.09.021
Received 31 July 2017; Received in revised form 11 September 2017; Accepted 12 September 2017
Available online 20 September 2017
0263-2241/ © 2017 Elsevier Ltd. All rights reserved.
 Measurement 114 (2018) 208–217

Nomenclature n is the number of parts

r is the number of trials
D is the basic major diameter of internal thread (nominal GRR is the repeatability and reproducibility
diameter) PV is the part variability, part variance
d is the basic major diameter of external thread (nominal TV is the total variability, total variance
diameter) ndc is the number of data categories, number of distinct ca-
D2 is the basic pitch diameter of internal thread tegories
d2 is the basic pitch diameter of external thread VB is the width of the flank wear land
D1 is the basic minor diameter of internal thread VBcr is the critical value of VB
d1 is the basic minor diameter of external thread KE is the shortening of the cutting wedge due to flank wear
H is the height of fundamental triangle ls is the spiral cutting length
P is the pitch of the thread UCL is the upper control limit
R is the radius of the curvature of the external thread con- LCL is the lower control limit
tour bottom ei is the error from zero line to inferior size of shaft
d3 is the internal diameter of the external thread with a es is the error from zero line to superior size of shaft
rounded root bottom P′ is the distance from the crest of one thread to the next,
EV is the repeatability, equipment variation P′ + P″ = P
R is the average range P″ is the distance from the left to right side of the crest,
AV is the reproducibility, appraiser variation P′ + P″ = P
XDIFF is the mean of differences, maximum average appraiser
difference

Fig. 1. Information flow in a system of technological quality assurance of a

metric screw thread (a) conventional system, (b) cyber-physical system.

factors.
- Stage 3, tolerance design, involves setting exact tolerances for those thread parameters or thread manufacturing indices whose nominal value
variations have a significant effect on the final variation.

The activities described above are related with the system design stage, while the technological quality of the metric screw thread is assessed with
regard to fulfilling geometrical requirements during the manufacturing process. It is conducted with on-line and off-line methods. The article
discusses the possibility of providing technological quality on the example of a metric screw thread. The authors elaborate on the limitations of the
thread measurement process and the possibility of quality providing in production development and manufacturing.

2. Materials and methods

In the case of ISO general purpose metric screw thread production, which is the most commonly applied thread in machine construction and
technology, the surfaces of the thread are described with parameters defined in ISO 68-1:1998 [17] (Fig. 2), which are subsequently tolerated into
the material, assuming as the zero line the nominal contour of the thread. A set of tolerance series and tolerance fields of the tolerated diameters is
shown in Table 1.
Developing thread specifications includes choosing a set of requirements regarding the geometrical parameters; e.g., for M8×1 6 h:

- type of thread: external, triangle, roller, right-hand 1-n,

209
 Measurement 114 (2018) 208–217

Fig. 2. Nominal values of the metric screw threads acc. to ISO 68-
1:1998.

- thread pitch: P = 1 mm,

- internal thread diameter: d = 8 mm,
- thread pitch diameter: d2 = 7.350 mm,
- thread contour angle: 60°,
- root bottom: curved,
- contour technology: turned thread,
- condition of roll surfaces: no burrs,
- type of thread end: with and without thread relief,
- manufacturing tolerance: 6 h.

The developed metric screw thread specifications can be implemented by use of various manufacturing techniques, both with weight-loss and
additive methods. The concept of manufacturing process relates to both geometrical accuracy of the thread and invariability of the created geo-
metrical quality in time. Therefore, technological parameters are defined depending on their purpose, quantity nature of production and manu-
facturing accuracy.
The concept of the manufacturing process assumed in M8×1 6 h metric screw thread specifications includes a set of requirements regarding:

- machined object: external thread diameter is ready-made, and will not be machined again, machined material is EN 41Cr4 steel,
- tools: cutting edge with universal metric screw thread shape with 1 mm pitch,
- machining tool: universal lathe with thread turning option by use of kinematic coupling method,
- manufacturing methods: thread turning by use of kinematic coupling method,
- feed: feed equal to thread pitch,
- infeed: radial plunge, used for smaller thread leads and strongly hardening materials, was conducted in several passes, as recommended: 5 passes,
infeed depth per pass: 0.20, 0.15, 0.13, 0.10, 0.08 mm,
- machining speed: as recommended by the tool producer: triangular one-side insert made of coated cemented carbide −vc ∈ (80,160) m/min.

3. Theory and calculations

3.1. Testing the geometrical accuracy of the thread during the manufacturing process

On the basis of preliminary tests, it was possible to proceed to the second stage of metric screw thread design – parameter design. This stage is
strictly connected with the use of experimental methods. The purpose is searching for nominal values for controllable factors with the lowest
sensitivity interference effects.
Tests were conducted for a given concept of threading. The metric screw thread created from thread turning was characterised by good visual
quality. There were no observed breaks in the thread, or surface notches. For this reason, it was assumed that the required thread quality had been
attained.
The geometrical accuracy of the thread had to be confirmed with thread measurements. As per industrial practice, it is most common to perform

Table 1
Tolerable metric screw thread diameters ISO 965-1:1998 [18].

Tolerable diameter Series of tolerances Location of tolerance field

d2 3, 4, 5, 6, 7, 8, 9 e, f, g, h
d 4, 6, 8
D2 4, 5, 6, 7, 8 E, F, G, H
D1 4, 5, 6, 7, 8

210
 Measurement 114 (2018) 208–217

measurements of external, internal, and pitch diameter of the thread. The measurement system is chosen depending on the tested tolerance T : for the
manufacturing conditions, it is assumed that the effective resolution for the measuring instrument is 0.1 T [mm], while for laboratory conditions, it is
bigger – 0.2 T [mm].
In the case of fine threads, all thread parameters are measured. External and internal, as well as roller and cone thread measurements can be
performed by use of measuring microscope, the three-wire method, or coordinate-measuring machines [11]. The application of optic-tactile probes
in coordinate-measuring machines enables one to perform measurements with error detection of about 0.35 μm [20,22].
Methods used in pitch (lead) measurements of metric screw threads are optic and contact measurement techniques. The essential advantage of
the optic measurements is a non-contact measurement method, a possibility to measure very small threads, convenient observation of the mea-
surement process on the screen, and high speed of measurements.
The most accurate measurement methods of screw thread pitch deviations are performed by a laser interferometer cooperating with an indexer,
while the nominal value of the pitch is usually determined fast and effectively by comparing the thread with a flat model of its shape. Non-contact
(optic) measurements are performed by use of a measuring microscope with expanded uncertainty at the 0.95 confidence level of about 1–2 μm. For
example, according to authors [11], in 3D object measurements, just as it is in the case of threads, expanded uncertainty (k = 2) of pitch mea-
surements by use of ZKM 01-250 CM and KKR microscopes is ± 1.8 μm. The repeatability of the microscope (2σ) at dynamic setting during the
measurements performed by use of measuring technique with bottom illumination is ± 2 μm at 95% confidence level.
Thread measurements on the coordinate-measuring machine are performed in relation to fine threads by means of a pin with diameters de-
pending on P pitch values of the measured thread. The calibration method of threaded models on the coordinate-measuring machine is described in
[3]. 2D scanning was conducted, and, on the basis of set of points, the uncertainty budget was calculated. In [13], the authors presented a mea-
surement uncertainty budget for a model thread pitch diameter, and performed an analysis of the influence of individual factors on the measured
value. An interesting solution in scope of thread pitch diameter testing is its estimation by use of points cloud from the CMM [21].
From the perspective of the manufacturing process, the chosen technology should make it possible to attain products within defined tolerance.
The following cases can be distinguished (Fig. 3):

- The system cannot manufacture a metric screw thread within the assumed tolerance (The system is not capable. The results are always beyond
tolerance.),
- The manufacturing system will always manufacture a metric screw thread within assumed tolerance (The system is capable. The results are
always within tolerances. The inspection system is not required.),
- The system is able to manufacture a metric screw thread within the assumed tolerance, but there is no certainty of achieving this goal (The system
is capable. The results are within tolerances. The inspection system is required.).

The values describing the metric screw thread were measured by use of direct methods or indirect values, for which a relation with the di-
mensional and shape accuracy was confirmed.
In order to be sure that the product has been manufactured in accordance with its specifications, and to close CLM, one should have a mea-
surement system able to perform quality control of the manufactured product. It is possible to use MSA methods (Measurement System Analysis) and
VDA (Verband der Automobilindustrie) [12,24]. The capability assessment of the measurement system should be conducted for each measurement
system, and should include values and value ranges for which the measurement system has the measuring capability (Table 2).
The device used for the measurements included machine vision using a digital camera for assessing metric screw thread quality, optical system
lens, illumination system, computer with image activation card, and software for digital image processing, identification and measurement of chosen
dimensional and shape parameters [25]. Fixing and positioning the 8 mm roller was performed in the centres. The authors used an illumination
system with a large workshop microscope light bulb for the transmitted light, and a reflected light LED illuminator. The image registered on the
camera's CCD was given in pixels. The images were saved in the camera's maximum resolution in 24-bit RGB model.
Depending on the CCD's resolution and the system's distance from the examined object, the result at 704 × 576 image resolution was 1 pix-
el = 0.00714 mm. The measurements were repeated 20 times with an incremental model for consecutive image resolutions by 2 operators. Loads,
linearity, repeatability, and reproducibility were checked according to the measurement systems MSA analysis [12], and the measurement system
capability within the scope of length measurements in tolerance was tested.

3.2. Measuring capability analysis

In order to fulfil specifications requirements of the thread, one should apply the CLM system (Closed Loop Manufacturing) for monitoring and
providing feedback with the designing and production developing system. In this case, it is necessary to perform measurements of the values defined
in the specifications by use of the measurement system of the measuring capability allowing for the assumed manufacturing tolerance [1,8].
Additional requirements may be provided for the measurement system, such as measuring time, or non-destructive testing. CLM loop will be closed
once the measurements have been performed and used for confirming the geometrical quality of the thread.
The calibrated vision system enables an automatic identification of the chosen geometrical properties of the object. Assessing the suitability of a
particular measurement system for monitoring the manufacturing process assumes fulfilling the capability conditions of the measuring instrument cg

Fig. 3. Application of measurement systems for providing technological

quality of the product.

211
 Measurement 114 (2018) 208–217

Table 2
The application of MSA to measurement system capability assessment.

Influence factors MSA 4 Company guidelines

Resolution/Data ndc ≥ 5 %RE ≤ 5% TOL
category
Repeatability of Systematic error must be small Cg ≥ 1.33
reference
Bias t-Test Cgk ≥ 1.33
Linearity t-Test %LIN ≥ 5%

Results for a linear gauge

Resolution/Data ndc ≥ 5 ndc = 16.32
category
Repeatability of Systematic error must be small GRR = 3.6 µm
reference
Bias t-Test – 95% confidence level for Upper Lower
load (zero in the confidence level) −4 µm 8 µm
Linearity t-Test p = 0.46

For GRR (gauge R & R) – an estimate of the combined variation of repeatability and reproducibility of the measurement system, ndc – number of distinct categories, Cg ,
Cgk – capability index, p – probability.

at defined tolerance value of the product, and appropriate small measuring uncertainty. According to the recommendations given by authors
[8,12,15], the value of the cg index should fulfil the condition:
0,2·T
cg = ⩾ 1,33
6·s (1)
and dependence with regard to the measuring uncertainty should be fulfilled.
2·s ⩽ 0,05·T (2)
Standard deviation s relates to the multiple measurements of the product taken from the process whose tolerance is T . In order to check the
suitability of the designed and manufactured vision measurement system for thread geometry measurements with a CCD camera, measuring
computer, illumination system in reflected and transmitted light, and software made in MATLAB, the authors performed measurements of a model
M8×1 screw. The property measured from the image was the M8×1 thread's pitch and pitch diameter (Td = 112 μm and Td2 = 180 μm of the screw
made within 6 h tolerance) in 12 section by 2 operators (see measuring results Fig. 4a and b, as well as Table 3).
The attained standard deviation s value for the M8×1 thread's pitch in reflected and transmitted light in the acquired image's resolution of
704 × 576 image points amounted to max 3 μm. The designated value of the capability index of the cg measuring instrument, based on formula No. 1
and 2, in consideration of the M8×1 tolerance (Td = −112 μm and Td2 = −180 μm of a screw made in 6 h tolerance), assumed basing on the
products' geometry specifications, amounted to over 1.33.
In order to determine the uncertainty of the whole measurement system, a procedure for uncertainty variation was used, which comprises of
repeatability of the measurement sys system EV (Equipment Variation), reproducibility of the measurement system AV (Appraiser Variation) and total
spread R & R (eng. Repeatability and Reproducibility) often marked as GRR [12,15]. The measurement system's capability assessment was conducted
for several cases, in which the variable was the type of applied illumination in the acquired image's resolution of 704 × 576 image points, at pixel
size amounting to 0.00714 mm.
The MSA guidebook evaluates the measurement system capability by limiting the calculated value of the model's repeatability (GRR ) with the
required assessment's criteria. The MSA guidebook recommends checking if the systematic error is sufficiently small. Moreover, the MSA procedure is
followed during the assessment of the number of data categories ndc . This number needs to be higher than 5. Exceeding the value for over the
criterion (specific limit) does not unequivocally point at the source of the excess. For this reason, the measurement system is usually corrected, and
the %GRR value calculated again whenever it happens to be below the assessment criterion.
The assessment criterion divides the measurement system capability into following categories:

- when %GRR < 10% – this range is most preferable, while it means that the system is unconditionally able to perform measurements of a given
value in the given tolerance,
- when 10% ≤ %GRR < 30% – the range can be, under certain conditions, accepted, depending on its applications,

Fig. 4. A box-plot diagram showing the values of the M8×1

thread's pitch [in mm] in reflected (a) and transmitted light
(b) for two operators (A and B) in image resolution of
704 × 576.

212
 Measurement 114 (2018) 208–217

Table 3
The uncertainty analysis of the whole measurement system during the measurement of the thread's pitch.

Value Calculation formula Measurement in Measurement in

reflected light transmitted light

Number of appraisers k = 2
R [mm] 1
(R + R2), k = 2 0.0053 0.0044
k 1
XDiff [mm] XDiff = Xmax −Xmin 0.0016 0.0009
EV [mm] EV = K1·R , K1 = 0.88 0.00072 0.00069
AV [mm] EV2
AV < 0 ⇒ AV AV < 0 ⇒ AV
AV = (XDiff ·K2)2−⎛ ⎞
=0 =0
⎝ n·r ⎠
K2 = 0.7071
GRR [mm] GRR = EV2 + AV2 0.0072 0.00069

M8×1 6h thread
d : ei = −180 µm, es = 0
TV [mm] TV = (UCL−LSL)/6 0.0300 0.0300
PV [mm] PV = TV2−GRR2 0.0296 0.0297
%EV [%] % EV = 100 (EV / TV ) 15.68 12.89
%AV [%] % AV = 100 (AV / TV ) 0 0
%GRR [%] % GRR = 100 (GRR/TV ) 15.68 12.89
%PV [%] % PV = 100 (PV /TV ) 98.76 99.16
ndc ndc = 1.41(PV / GRR) 8.87 10.84

d2 : ei = −112 µm, es = 0
TV [mm] TV = (UCL−LSL)/6 0.0187 0.0187
PV [mm] PV = TV2−GRR2 0.0180 0.0183
%EV [%] % EV = 100 (EV / TV ) 25.20 20.73
%AV [%] % AV = 100 (AV / TV ) 0 0
%GRR [%] % GRR = 100 (GRR/TV ) 25.20 20.73
%PV [%] % PV = 100 (PV /TV ) 96.77 97.83
ndc ndc = 1.41(PV / GRR) 5.41 6.65

- when %GRR ≥ 30% – the system cannot be accepted in given application.

According to the presented data tables, it is possible to conclude that the optical measurement system can be used for measuring M8×1 metric
screw thread in tolerance 6 h, because the system fulfils all assessment criteria.
In the case of the applied universal instrument, where the external diameter of the thread had been ready-made earlier, the ndc and %GRR
indices, both in reflected and passing light, fall within the assumed variability range. Had an incorrect thread manufacturing technology been
applied, there could have appeared burrs, or other interferences (breaks in the thread, surface corrugation, accretion, wear), which would have
caused deterioration of the external diameter's accuracy. The proposed solutions for the future are ones which would include developing tech-
nologies for maintaining the accuracy of the external diameter of the thread.
In the case of pitch diameter, 6 h tolerance, ei equal to −112 µm is smaller than ei for the external diameter. Maintaining tolerance during thread
production results from the turning process and limiting confounding factors, such as vibration or tool wear. Measuring capability criteria for the
optical system are observed, but in the consequence of the lower ei value, the key role, in this case, is performing measurements by a professional
operator, or developing an automatic reading system of the measured value.

4. Results and discussion

4.1. Determining controllable and uncontrollable factors

The geometrical accuracy of threads achieved in the machining process is characterised by strong dependence on initial conditions and sustaining
those conditions throughout the process. It is a particularly difficult task, especially during high speed cutting, or in case of hard-workable materials
[1,8].
Controllable factors in the scope of geometrical accuracy of the thread are those factors which provoke systematic constant or varying geo-
metrical deviations, which may be reduced during the machining process. Constant systematic deviations involve deviations caused by geometrical
or kinetic inaccuracy of the machine tool, wrong machine tool setting or geometrical inaccuracy of the tool itself. Varying systematic deviations
consist in those deviations which are caused by events taking place during the machining process on the regular basis. They involve deviations
arising as a result of tool wear or machine warmth deformations.

• The studies presented in [5] suggest that machining conditions, especially amplified temperature in the machining area, result in higher values of
tensile stresses in the core of the screw. It has been evaluated that tensile stresses increase when machining parameters surpass the optimal scope,
or in the case of cutting edge’s wear. In order to maintain machining conditions, it is possible to model the influence of the parameters of the
process on the machining force, and then to control their values during machining.
• In [23] the authors describe a method used for choosing the coating of the tool for threading a titanium alloy, induced by ultrasonic vibrations.
Experimental research on determining the factors influencing the quality of thread shaping, as well as technological recommendations for
maintaining the machining force on a given level are presented in [6,9].

213
 Measurement 114 (2018) 208–217

• In the machining process, it is possible to determine the minimum synchronisation error of the axis movement with the rotation of the spindle,
while modelling the reciprocal relation of the tool and the machined object. The research presented in [2] consisted in determining the influence
of the torque on the process of threading. It has been determined that the synchronisation error should be maintained on the 0.04 mm level, that
is 8% of tap's lead, and then it is possible to attain class 2 accuracy of JIS thread (6H ).

Uncontrollable factors in the scope of geometrical accuracy of the thread are those factors which are difficult to predict, and result from
unwanted events. They involve deviations caused by heterogeneous hardness of the machined material and its structure, heterogeneous properties of
the tool's material, diverse machining allowances, the system's vibrations, etc.
Recognising controllable and uncontrollable factors in the thread machining process, and their subsequent analysis enables one to determine
quality relations. This is because current thread geometrical accuracy depends on forcing factors, machined material properties and tool properties,
as well as other factors [28]. There are three basic events affecting the accuracy of threading in the machining process:

- occurrence of reversible processes of various intensity, e.g. instability of the machining system,
- discrete change of one of the geometrical parameters of the thread beyond the acceptable limits, e.g. in the effect of catastrophic failure of the
tool,
- the effect of slow, irreversible degradation processes (e.g. tool wear).

Technological quality of the thread, created in the machining process, is the result of the variable during the machining force, which can bring
about negative effects, e.g. render itself on the machined surface as the result of dynamics change, as well as affect dimensional and shape accuracy.
Stability analysis of the system has been performed by means of a seismic sensor (with sampling frequency of – 44100 Hz, sampling resolution –
16 bit) for the full scope of rotational speeds and machined layer section in the range of 0.05–0.2 mm2/rev. It has been determined that whenever the
cutting edge falls within the admissible wear values, the system is stable for all examined measuring points. Fig. 5 shows two spectrograms depicting
vibrations signal of the machining system in the primary machining period in the case of a new cutting edge and one with certain degree of wear
(VB < VBcr ). The occurring vibrations are quickly dampened, there are no self-generating vibrations.

4.2. Providing the accuracy of the thread during the machining process

There are many methods to check if threading accuracy is provided during the machining process. They include both on-line and off-line
methods, methods providing a direct measurement of the metric screw thread geometry, as well as indirect methods [11]. One of the methods of
threading accuracy assessment is the optical method, consisting in applying a visual system with an automatic view analysis to determine mea-
surement characteristics of the thread. In [14] the authors describe a way of assessing thread quality by optical method by with the help of a digital
camera. The results were compared with measurements performed with a workshop microscope. In [25], on the other hand, the authors presented a
microscope measurement system for assessing the dimensional and shape accuracy of threads; the system’s quality has been compared with the
results obtained for a coordinate-measuring machine and a laser scanner. The imaging methods for thread images analysis have been compared in
many papers, e.g. [16] or [25]. The parameters most often chosen for assessment are geometrical thread parameters, such as pitch, contour height, or
thread angle.
The inspection method of internal threads for car industry has been presented in [7], where the authors analysed internal threads used in this
branch. It has been determined that most of them are of secondary importance. However, there are threads which are identified as having a
significant impact on the subsystem’s behaviour, or e.g. even the engine’s performance. In such case, most often applied methods are tactile methods
or CMM [19,20,22].
In order to fulfil the requirements of the inspection time and measuring accuracy, the authors have developed in [7] a method of assessing the
quality of the internal thread. The system comprised of two complementary devices: a laser scanner, which was displaced with help of a stepper
motor, and machine vision – a linear scanner connected to a camera with an optical “sight pipe” system.
The vision system (VS) applied in the threading accuracy tests during the machining process was built at Koszalin University of Technology as a
modification of a stationary system – a microscopic measuring system [25]. The underlying assumptions for the vision system, integrated as an
immanent element of the manufacturing system, have been presented in [26]. The modification of the vision system for thread measurements
consists in integrating a microscopic measuring system into the machining system and performing measurements during the machining process. The
developed vision system made it possible to perform measurements after consecutive infeeds of the tool. The image was registered with a digital
camera IKEGAMI ICD-47E with a CCD 1/2“ transducer with an effective resolution of 752 (H) × 582 (V) pixels. The camera was connected to a
computer. The vision system applied in the measurements made it possible to observe the object in transmitted and reflected light (Table 4).

Fig. 5. Spectrograms of vibrations signals for TNMG

160408 PF 4015, velocity 200 m/min, machined cross sec-
tion 0.15 mm2/rev, (a) cutting edges in the primary wear
stage,(b) cutting edges with VB = 0.22 mm.

214
 Measurement 114 (2018) 208–217

Table 4
A comparison of digital images of the M8×1 thread contour during the turning process after three, six and nine infeed passes.

Fig. 6. A comparison of mean results of changes in

thread geometry (a) half of the pitch calculated for the
root, (b) half of the pitch calculated for the crest of the
thread.

Fig. 7. (a) Mean wear of the cutting insert’s cutting edge on

the auxiliary clearance surface (b) cutting edge reduction.

Table 5
A comparison of digital images of the cutting edge for three cutting distance values.

A usefulness assessment of the vision system in machining was performed for metric screw threads by measuring a M8 × 1 bolt with a roller
thread. The bolts were made from 35SG steel, which undergone the process of heat treatment to 30 HRC hardness. The metric screw threads were
produced with tolerance of 6h [18]. The full tool’s infeed was divided into nine passes. Measurements were performed in 12 sections in transmitted
and reflected light. Table 4 show exemplary images taken for illumination in transmitted light, as well as the result of computer imaging by use of

215
 Measurement 114 (2018) 208–217

Table 6
An analysis of the measurement capability assessment of the vision system for d and d2 , according to the MSA procedure, with consideration of the changes in the cutting
edge during the cutting process.

Value Calculation formula Reflected light Transmitted light

M8 × 1 6h thread + correction in the machining program

EV [mm] EV = R ·K1,GRR = EV2−AV2 0.0047 0.0048
AV [mm] EV2
0 0
AV = (XDIFF ·K2)2−⎛ ⎞,
⎝ n·r ⎠
K2 = 0,7071
GRR [mm] GRR = EV2−AV2 0.007 0.006
TV [mm] TV = (UCL−LSL)/6 0.0300 0.0300

For ls = 1000 m, KE = 0.0049 mm

d : ei = −180 µm + KE (ls ) , es = 0
TV [mm] TV = (UCL−LSL + KE (ls ))/6 0.0291 0.0291
PV [mm] PV = TV2−GRR2 0.0288 0.0289
%EV [%] % EV = 100 (EV / TV ) 16.12 13.25
%AV [%] % AV = 100 (AV / TV ) 0 0
%GRR [%] % GRR = 100 (GRR/TV ) 16.12 13.25
%PV [%] % PV = 100 (PV /TV ) 98.69 99.11
ndc [] ndc = 1.41(PV / GRR) 8.63 10.54

d2 : ei = −112+KE (ls ),es = 0

TV [mm] TV = (UCL−LSL + KE (ls ))/6 0.01785 0.01785
PV [mm] PV = TV2−GRR2 0.0172 0.0174
%EV [%] % EV = 100 (EV / TV ) 26.36 21.67
%AV [%] % AV = 100 (AV / TV ) 0 0
%GRR [%] % GRR = 100 (GRR/TV ) 26.36 21.67
%PV [%] % PV = 100 (PV /TV ) 96.43 97.62
ndc [] ndc = 1.41 (PV / GRR) 5.15 6.34

logarithmic edge filter.

4.3. Determining the influence of the cutting edge’s wear on threading accuracy

In the case of thread shaping by turning, the most significant confounding factor affecting the dimensional and shape accuracy is wear of the
cutting edge. In order to perform an analysis of the cutting edge’s wear on the geometrical shape of the thread, the authors gathered data from the
image concerning the changes in the thread’s geometry taking place during the turning process due to the wear of the cutting insert’s cutting edge.
The values of “half the thread’s pitch” f for p′ roots and p″crests were measured on the pitch diameter. For statistical analysis values p′ and p″were
determined many times, in 45 thread measurements for each of the three measuring points (for 500 m, 1000 m, and 1500 m of cutting distance).
Spot graphs (Fig. 6) depict changes in thread geometry, in its crest and root, determined on the basis of digital images for a given cutting distance
(see Fig. 7).
The analysis of the charts shows clearly that longer cutting distances and shorter cutting edges results in changes in geometry of the thread. The
longer the cutting distance, the bigger the crest (beyond assumed tolerances), and the thread becomes less and less symmetrical. The results of the
changed geometry of the cutting edge obtained from images filtered with logarithmic edge filters and Laplace’s filters were comparable.
The most representative and most often assumed measure of the cutting edge’s wear is the wear width on the clearance surface (main or
auxiliary) – VB value. VB value was measured in such a way that each of the clearance surfaces (main or auxiliary) was submitted to five tests. The
value of VB cutting edge’s wear parameter was determined from a digital image, according to the procedure described in [27]. The measuring system
of wear values was able to perform wear measurements with ± 0.004 mm accuracy at 95% confidence level.
Table 5 shows exemplary digital images of a cutting edge from its auxiliary side of the clearance surface for three measuring points. The
measurement results of the VB index for the cutting edge’s wear, averaged for the auxiliary and mean clearance surfaces, have been used to
determine the KE cutting edge’s reduction index, according to the equation presented in Table 5. KE cutting edge’s reduction has been modelled in
relation to the size of wear and geometry of the cutting edge (γ rake angle and α clearance angle).
Fig. 3 shows mean VB wear of the cutting insert’s cutting edge (determined on the basis of a digital image) on the auxiliary clearance surface for a
cutting distance of 500 m, 1000 m and 1500 m, as well as KE cutting edge reduction change in cutting distance function, determined according to
the equation presented in Table 5.
The occurrence of a disturbance in the system, identified as provoking machining errors, is taken into consideration in the control system. A
system of this kind may perform a correction during the machining process or afterwards. For example, tool wear and cutting edge’s apex dis-
placement may be used for correcting its position. In this way, the thread will be manufactured within 6h tolerance, despite the possible disturbance
in form of cutting edge reduction.
Table 6 shows the results of the analysis of measurement capability of the vision system applied for providing the assumed measuring accuracy of
the M8 × 1 thread, in consideration of the procedure from the MSA handbook [12,18]. The required accuracy for manufacturing a 6h thread requires
one to take into account wear of the cutting edges, as even the admissible value of VB wear significantly affects final accuracy of the cutting edge’s
dimensions.
In order to perform capability assessment of the measuring system for measuring a given value in assumed tolerance, one should use two indices:
%GRR and ndc .
Index %GRR relates to repeatability and reproducibility of the measuring system, in consideration of tolerance of the measured part and total

216
 Measurement 114 (2018) 208–217

measuring system variability. If %GRR value is higher than 10% and lower than 30%, the assessment should be considered satisfactory; however, one
should conduct a periodical verification of the measuring capability of the system. Moreover, the indices make it possible to ascertain that the system
is not over dimensioned.
The value of the second ndc index, in the context of thread pitch assessment both on the thread diameter, as well as the external diameter, does
not fall below the admissible value of 5. However, one should note that ndc value, which, according to MSA procedure, represents the number of data
categories, is close to this value, for a vision system using reflected light. This means that whenever a higher measuring accuracy is needed, one
should apply illumination in form of transmitted light for thread contour observations.
Both in the case of vision system with light reflected from the surface and vision system with transmitted light where thread contour is observed,
the vision systems were able to perform measurements of the thread pitch on the thread diameter and the external diameter.

5. Conclusions

Cyber-physical system with virtual metrology is able to provide the tools for the product to be manufactured with assured accuracy. The concept
of the cyber-physical system allows for the product quality requirements the transformation into the cybernetic world of the measurement model and
manufacturing model. In this way the measurement model becomes part of the overall manufacturing system.
For the development of the measurement model, an optical measurement system has been developed to measure the characteristics of the
workpiece on the basis of the subject's image in passing and reflected light. The measuring system was analyzed in the M8×1 thread measurement
and its measurement capability was checked at the same time.
The model of the manufacturing system was elaborated and the influence of factors affecting the technological quality assurance of the product
was analyzed. A more thorough analysis was made to assess the effect of tool flank wear on the accuracy of the metric screw thread in the
manufacturing process model. At the same time, capacity analysis was performed for the entire manufacturing system.

References

[1] S. Adamczak, W. Makieła, Analyzing variations in roundness profile parameters during the wavelet decomposition process using the Matlab environment, Metrol. Measur. Syst. 18
(1) (2011) 25–34.
[2] J.H. Ahn, et al., Effects of synchronizing errors on cutting performance in ultra-high tapping, CIRP Ann.-Manuf. Technol. 52 (1) (2003) 53–56.
[3] S. Carmignato, L. de Chiffre, A new method for thread calibration on coordinate measuring machines, CIRP ANN. 52 (1) (2003) 447–450.
[4] M. Diering, K. Dyczkowski, A. Hamrol, New method for assessment of raters agreement based on fuzzy similarity, in: 10th International Conference on Soft Computing Models in
Industrial and Environmental Applications, Part of the Advances in Intelligent Systems and Computing, AISC 368, 2015, pp. 415–425.
[5] E. Fetullazade, H.K. Akyildiz, S. Saritas, Effects of the machining conditions on the strain hardening and the residual stresses at the roots of screw threads, Mater. Des. 31 (2010)
2025–2031.
[6] R.V. Guseinov, M.R. Rustamova, Internal thread cutting in the presence of radial forces, Russ. Eng. Res. 29 (5) (2009) 480–481.
[7] E. HONG, H. ZHANG, R. KATZ, J.S. AGAPIOU, Non-contact inspection of internal threads of machined parts, Int. J. Adv. Manuf. Technol. 62 (2012) 221–229.
[8] Z. Humienny, M. Berta, New multimedia geometrical tolerancing course, in: Proceedings of 12th CIRP Conference on Computer Aided Tolerancing, Electronic Proceedings,
Huddersfield, 2012.
[9] V. Ivanina, Influence of parameters of the cutting part of taps on threading accuracy, Meas. Tech. 48 (10) (2005).
[10] J.T. Krishankant, B. Mohit, K. Rajesh, Application of Taguchi Method for Optimizing Turning Process by the effects of Machining Parameters, Int. J. Eng. Adv. Technol. (IJEAT) 2 (1)
(2012) 263–274.
[11] W. Jakubiec, W. Płowucha, M. Starczak, Analytical estimation of coordinate measurement uncertainty, Measurement 45 (10) (2012) 2299–2308.
[12] Measurement Systems Analysis Reference Manual, Fourth ed. AIAG, Michigan, USA, 2010.
[13] T.P. Merkac, B. Acko, Comparising measuring methods of pitch diameter of thread gauges and analysis of influences on the measurement results, Measurement 43 (3) (2010)
421–425.
[14] J. Mutambi, L. Yu, Application of digital image analysis method in metric screw thread metrology, J. Shanghai Univ. (English Edition) 2 (2004) 208–212.
[15] P. Pajzderski, M. Diering, The % R & R index-Proposals and guidelines on the reference value and acceptance criteria, J. Trends Develop. Machin. Assoc. Technol. 16 (1) (2012)
111–114.
[16] E. Peng, G. Zhang, Image processing technology research of on-line thread processing, Energy Procedia 17 (2012) 1408–1415.
[17] ISO 68-1: 1998, ISO general purpose screw threads — Basic profile — Part 1: Metric screw threads,1998.
[18] ISO 965-1:1998, ISO general purpose metric screw threads —Tolerances — Part 1: Principles and basic data, 1998.
[19] M. Poniatowska, Deviation model based method of planning accuracy inspection of free-form surfaces using CMMs, Measurement 45 (5) (2012) 927–937.
[20] T. Kowaluk, E. Ratajczyk, Analysis of the Influence of the Type of Mounting of the Coordinate Measuring Arm to its Accuracy, University of Bielsko-Biała Press, 2014, pp. 82–86.
[21] I.A. Shchurov, Calculation of the virtual pitch thread diameter using the cloud of points from CMM, Int. J. Adv. Manuf. Technol. 53 (1 4) (2011) 241–246.
[22] J. SLADEK, Coordinate Metrology: Accuracy of Systems and Measurements, Springer Verlag GmbH, Berlin Heidelberg, 2016.
[23] C. Tsao, K. Kuo, Ultrasonic-assisted vibration tapping using taps with different coatings, Trans. Nonferrous Met. Soc. Chin. 22 (2012) 764–768.
[24] VDA 5, Capability of Measurement Processes, second ed., VDA, Berlin, Germany, 2011.
[25] B. Storch, I. Wierucka, Optical measurements of repeatable contours using the image processing techniques, Acta mechanica et automatica 1 (2) (2007) 59–62.
[26] A. Zawada Tomkiewicz, D. Tomkiewicz, Mechatronic monitoring tool (MMT) for machined surface evaluation during the cutting process – (ID 269) ISMQC 2013, in: 11th
International Symposium on Measurement and Quality Control, 11-13 September 2013, Krakow, 2013.
[27] A. Zawada-Tomkiewicz, Measurement uncertainty assessment in machine vision system for tool wear estimation, Coord. Meas. Techn. (2008) 175–183.
[28] A. Zawada-Tomkiewicz, Theoretical and experimental basis for monitoring the process of turning, using information on the stereometric features of the machined surface, Press of
Koszalin University of Technology, Koszalin, 2012.

⁎
Anna Zawada-Tomkiewicz , Izabela Wierucka
Koszalin University of Technology, 75-620 Koszalin, 15-17 Raclawicka Str., Poland
E-mail address: [email protected]

⁎
Corresponding author.

217