An Analytical Model for Pitch

Moment Stiffness of Bolted

Akshay D. Harlalka1
Connections and Its Application
Department of Mechanical Engineering,
Massachusetts Institute of Technology,
in Ballscrew Bearing Support
77 Massachusetts Ave,
Cambridge, MA 02139 Block Selection

Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022

e-mail: [email protected]
Bolted joints are commonly used structural connections as they provide a strong secure
Alexander H. Slocum joint along with ease of assembly/disassembly. While analytical models for the axial stiff-
Department of Mechanical Engineering, ness of bolted joints are well developed, models for moment (angular) stiffness of bolted
Massachusetts Institute of Technology, structures, such as ball screw bearing support blocks, are needed to help engineers
77 Massachusetts Ave, rapidly design more efficient precision machines. This paper develops a parametric
Cambridge, MA 02139 moment stiffness model for bolted connections which is verified via numerical and experi-
e-mail: [email protected] mental methods. Application of the model is illustrated with a ball screw system design
spreadsheet, available in Supplemental Material on the ASME Digital Collection,
applied to two case studies (machine tool linear axis and high-speed 3D printer) to show
how predicting the moment stiffness of ball screw support bearing blocks helps in expanding
the available design space and enhance the design performance. [DOI: 10.1115/1.4054474]

Keywords: ballscrews, bolted connections, moment stiffness, pillow blocks, bolted joints,
design of machine elements, design theory

1 Introduction determined the radii of separation of the bolted members using

experimental and finite element methods which were also in agree-
The computation of bolted connection stiffness has been studied
ment with FEA. Studies have found that in all cases, variables such
since the early 1970s with a focus on stiffness in axial and lateral
as clamping pressure distribution [8], and the changing contact con-
directions [1–6]. The stiffness ratio between the bolt and the
dition of the members under external loads [9], contribute to the
clamped members directly governs the fraction of the load that
stiffness of bolted connections.
will be safely carried by the bolt [3]. Zhang and Poirier [1] pre-
These theories, however, do not directly apply to the moment stiff-
sented a model for bolted joints that considers effects related to rota-
ness of bolted joint connections and FEA and experiments are gen-
tion of the members and their additional deformation when
erally too costly. So engineers often use rules-of-thumb or copy what
subjected to external loads and verified the model with finite
has been done before which may not lead to good results. In a
element analysis (FEA). Nassar and Abboud [2] proposed an
fast-paced engineering environment, analytical models are needed
improved stiffness model based on more accurate determination
for rapidly creating and developing viable design concepts. Once a
of the effective joint area where the strain cone angle was experi-
design has been selected and developed in this manner, FEA can
mentally determined and various effects related to joint sizes,
then be economically used to check the design and ensure an ade-
under head contact radii ratio, and plate material/thickness ratio
quate safety factor.
were investigated. The results of the proposed model agreed well
The moment stiffness of bolted connections is an important
with experimental data and FEA. Other studies have focused on
design parameter for engineers working on the design of precision
the computation of member stiffness. Lehnhoff et al. [4] calculated
machines. One such example is a design of ball screw drive system.
the member stiffness of various bolted connections (with different
The selection of a ball screw drive system for a given application is
bolt sizes, materials, and thicknesses). Dimensionless curves were
often an iterative process as shown in ball screw supplier catalogs
presented in the paper to help the audience estimate the member
[10]. Knowledge of the system stiffness helps develop servo
stiffness ratios for a variety of bolted joints. The analytical predic-
control models and predict positioning accuracy. The system stiff-
tions were in moderate agreement with FEA. Sethuraman and Sasi
ness is a function of the stiffness of the screw shaft, nut, support
Kumar [5] did a similar evaluation of member stiffness of axisym-
bearings, and housing which are strongly affected by bolted joint
metric bolted connections and proposed various empirical formulas
stiffness. Established guidelines are available to predict the stiffness
for computation of the same which were in agreement with the
of all these components except the moment stiffness of the housing.
FEA. The study of pressure distribution among bolted members
The theory presented in this paper can be used to predict the stiff-
has also been motivated by interests in other domains of electric
ness of bearing block housing thereby helping the designer to deter-
or thermal power across bolted connections. Gould and Mikic [7]
mine the full stiffness model as a part of the ball screw selection
studied the pressure distribution in the contact zones and
procedure.
An analytical model is presented herein for the moment stiffness
1
Corresponding author.
of a bolted connection and is verified via finite element analysis and
Manuscript received November 13, 2021; final manuscript received April 25, 2022; experimental testing. Three different ball bearing mounting blocks
published online July 19, 2022. Assoc. Editor: Joshua Summers. (i.e., pillow blocks) are used as examples. For each pillow block,

ASME Open Journal of Engineering 2022, Vol. 1 / 011027-1

Copyright © 2022 by ASME
analytical calculations, FEA solutions, and experimental testing strain cone with a half-apex angle of α is represented in Figs. 4(a)
results are compared and discussed. and 4(b), respectively. The shape of the pressure distribution
Herein, Secs. 2–4 develop the stiffness analysis of bolted pillow among the clamped members of a bolted connection is complex,
blocks and discuss the generic FEA analysis procedure, and detail and Shigley [11] suggests the use of simpler cone geometry with
the experimental setup. Sections 5–10 present the results of the a half-apex angle of 30 deg to approximate the shape of the strain
study for three types of pillow blocks. Opportunities for further cone. For members with unequal thickness and materials, the calcu-
work observed during this study are presented in the Conclusion. lation of the member stiffness becomes cumbersome. As per
Shigley [11] and Williams [12], for the case of members with
unequal thicknesses, the strain cone extends up to half of the effec-
2 Development of Analytical Model for Pitch Moment tive grip length (p) of the clamped members and then begins to
Stiffness of Bolted Connections recede for the other half of the grip length. Generally, it is desirable
to make the strain cones of the bolted connections overlap in order
2.1 Sources of Compliance. Consider a steel pillow block to prevent inducing any straightness errors on the parts being
bolted onto an aluminum base (Fig. 1) and a model of the stiffness bolted [13].
along the pillow block axis (Fig. 2). Consider the base to be fixed,

Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022

This theory can be visualized with photoelastic experiments of
there are two main sources of compliance in the system: one related plexiglass members of unequal thicknesses. To observe the stress
to the compliance of the pillow block itself (shear, bending, and field, the bolted assembly was kept directly in front of a polarized
torsion) and the other related to the bolted joint stiffness. Since light source (a liquid crystal display (LCD) screen in this case). A
the focus of this paper is on prediction of the joint stiffness, a circular polarizer filter (55 mm diameter) was rotated accordingly
pillow block with much greater inherent stiffness than its bolted to block the light from the LCD screen and to observe the stress
connection was preferred for study, so the system stiffness will be fields from the plexiglass members clearly. Figures 5(a) and 5(b)
dominated by joint stiffness. show these stress fields in the clamped members. Stress birefrin-
gence is just one of the many methods to observe strain cones
2.2 Different Cases. To investigate the efficacy of the analyt- and contact pressure distribution in interfaces. Researchers have
ical model, pitch moment stiffnesses (“pitch stiffness”) of three dif- also used ultrasonic measurements [14] and pressure film sensors
ferent types of pillow blocks were assessed with the model and [15] to investigate the distribution of contact pressure and predict
compared with the experimental data and numerical solution. The interfacial contact stiffnesses.
pillow blocks include 2-bolt type with steel body, 4-bolt type a. Member stiffness calculation:
with steel body, and 2-bolt type with plastic polylactic acid Traditional model
(PLA) body (Fig. 3). For experimental measurements, the predicted
pitch stiffness is transformed into a linear transverse stiffness paral- From Fig. 4(b), it can be seen that the flange consists of two frus-
lel to the direction of pillow block axis. While the load is applied tums and the bed consists of a single frustum. To compute the
along the axis of the pillow block and the deflection is measured member stiffness of the flange, the compliances of the two frustums
at the tip of the block as shown in Fig. 1. This is also useful to need to be calculated separately and then added together using
the machine designer when considering the stiffness of a mounted Eq. (8). As per Shigley [11], Eq. (1) can be used to calculate the
ball screw for example. compliance of a frustum under the head of the bolt.


(2h1 tan α + D − d)(D + d)
2.3 Nomenclature. Table 1 represents various symbols used ln
in the development of the analytical model and how they relate to (2h1 tan α + D + d)(D − d)
Ctop|frustum = (1)
the bolted pillow block assembly (Fig. 1). πEflange d tan α

It can be observed from Eq. (1) that the compliance of the frustum
2.4 Joint Stiffness: Pitch depends on its height, half-apex angle of the cone, bolt major and
2.4.1 Axial Joint Stiffness Model. The axial stiffness of the head diameters, and the elastic modulus of the material. The
joint is governed by the stiffness of different members in the struc- height of top and bottom frustums of the flange (h1 and h2) can
tural loop, including the bolt stiffness and the member stiffness. The be derived using simple geometry from Fig. 4(b) (refer Eqs. (3)
traditional spring model of the bolted joint and a visualization of the and (4)). A cone angle of 30 deg is used for the calculation. The

Fig. 1 Illustrations of a pillow block assembly mounted on a metal base: (a) pillow block assembly and (b) dimensional
representations

011027-2 / Vol. 1, 2022 Transactions of the ASME

 Fig. 2 Spring model of the pillow block mounted on a metal base

Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022

Fig. 3 Different test cases: (a) 2-bolt pillow block (steel), (b) 4-bolt pillow block (steel), and (c) 2-bolt pillow block (PLA)

flange (pillow block) is made of plain carbon steel (Eflange = detailed overview of the bolt parameters and the calculations,
200 GPa). The total effective grip length includes the height of please see Supplemental Material on the ASME Digital Collection.
the flange and the height of the base material in grip. Since, this
is a tapped joint, the equivalent height of the base material in grip p = hflange + hbg (2)
can be assumed to be equal to half the bolt diameter [11]. For the
h1 = p/2 = (hflange + hbg )/2 (3)
Table 1 Symbols used in the development of analytical model

Symbol Meaning h2 = p/2 − hbg (4)

W Width of the pillow block

H Total height of the pillow block hbg = d/2 (5)
Hc Height of the pillow block axis
Ht Depth at which threads start in the base
Equation (1) only considers the stiffness of the top frustum. A
L Length of the pillow block similar equation can be used to compute the stiffness of the
E Young’s modulus bottom frustum (Eq. (6)). However, to do so, two parameters will
Eflange Young’s modulus of pillow block (flange) need to be modified. The head diameter (D) will need to be replaced
Ebed Young’s modulus of base (bed) by the interfacial contact diameter (D′ ) and the thickness of the cone
I Moment of inertia frustum will need to be h2. The interfacial contact diameter can be
G Shear modulus calculated using simple geometry as per Eq. (7).
θ Angle of twist


J Polar moment of inertia (2h2 tan α + D′ − d)(D′ + d)
h1 Height of the top cone frustum of flange ln
Height of the bottom cone frustum of flange (2h2 tan α + D′ + d)(D′ − d)
h2 Cbottom|frustum = (6)
hbg Height of base material in grip πEflange d tan α
hflange Height of the pillow block (flange) at bolt underside
hbed Height of the bed
t Effective thickness for shear stiffness calculation D′ = D + 2hbg tan α (7)
d Bolt major diameter
D Bolt head diameter To find the equivalent flange compliance, the compliances of the
D′ Interfacial contact diameter top and bottom frustum are added in Eq. (8). For bed compliance
α Half-apex angle of the cone calculation (Eq. (9)), the height of the frustum is same as the
p Effective grip length of the clamped members height of the base material in grip. The equivalent head diameter
Ad Area of unthreaded portion of bolt (D) for the bed cone frustum can be assumed to be 1.5d [11],
At Area of threaded portion of bolt which is almost equal to the bolt head diameter (D).
ld Length of unthreaded portion of bolt in grip
lt Length of threaded portion of bolt in grip Cflange|comp = Ctop|frustum + Cbottom|frustum (8)
τ Shear stress
F Force


Cx Compliance of x (2hbg tan α + D − d)(D + d)
Kx Stiffness of x ln
Z Length of the side flange of pillow block measured from bolt (2hbg tan α + D + d)(D − d)
Cbed|comp = (9)
centerline to main body πEbed d tan α
Pext Pressure due to external loading
Ppreload Preload pressure In the traditional member stiffness model represented by
Fig. 4(a), the equivalent member stiffness of a bolted connection

ASME Open Journal of Engineering 2022, Vol. 1 / 011027-3

 Fig. 4 Bolted connection visualization and spring model: (a) conventional spring model of the
axial joint stiffness and (b) bolt visualization

Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022

can be calculated using Eq. (10). connection. The experimental value for stiffness of this member
was reported to be 565 kgf/um or 5540 N/um [16].
1
Kmember(Traditional) = (10) b. FEA results
Cflange|comp + Cbed|comp
Using the same geometry of the circular plate and pressure plates
Contribution of shear stiffness used in Maruyama’s experimental setup, the deformation of the
member plate was simulated with FEA. As the objective of the
While some bolted connection spring models [11] take just com- model was to probe the member stiffness only, the bolt shank,
pression stiffness of the members into consideration, other spring head, and nut were not modeled. The faces of the circular plate at
models take the member shear stiffness also into account [13]. the interface locations of the pressure plate were subdivided. All
Hence experimental data on member stiffness from earlier literature degrees-of-freedom at the bottom pressure plate interface were
were gathered and compared to results from FEA simulations and fixed, while for the top pressure plate interface, only X and Y
various hypothesized analytical models. Some of these analytical degrees-of-freedom were fixed. The X–Y constraint on the top inter-
models took shear stiffness into account while other models did face was important because the model assumed no occurrence of
not. An FEA model corresponding to a member stiffness experi- slip between the circular member plate and the pressure plates.
ment setup from the published literature [16] was created and its Based on the FEA result, the stiffness of the clamped plate was
results were compared with published experimental data for valida- reported to be 5555 N/um, which is very close to Maruyama’s
tion. After validation, FEA results were used to pick the accurate experimental result of 5540 N/um. For a visual representation of
analytical model for member stiffness. the loads and constraints applied during this analysis, please refer
Analytical, FEA, and experimental validation of member to Fig. 7(a).
stiffness
c. Analytical results
a. Experimental results
Maruyama [16] used a simple experimental setup to determine Using the member compliance formulas (Eqs. (1)–(10)) listed by
the axial stiffness of the members in a bolted connection. The Shigley [11], a theoretical prediction of the member stiffness can be
setup consisted of a circular plate with a hole and two pressure made. For the parameters corresponding to Maruyama’s experimen-
pieces, simulating bolt head, and nut to apply the axial load. The tal setup, the axial stiffness of the member plate is calculated to be
diameter of the hole was 25 mm and the diameter of the circular 6420 N/um.
plate was 100 mm. The material of the circular plate was S45C (des- d. Conclusion
ignated according to the Japanese Steel Grading System). Its Amer-
ican equivalent is AISI 1045. A single circular plate was used as a Based on the aforementioned values, the FEA, experimental, and
replacement for two clamped members found in a bolted theoretical calculations are in reasonable agreement with each other.

Fig. 5 Photoelastic experiments for stress field observation: (a) stress field observation with 10 mm and 3 mm thick plates
and (b) stress field observation with 50 mm and 10 mm thick plates

011027-4 / Vol. 1, 2022 Transactions of the ASME

 Fig. 6 Different stiffness models: Understanding the role of shear stiffness in a
bolted connection: (a) Model 1: compression stiffness only, (b) Model 2: com-
pression with shear stiffness parallel, and (c) Model 3: compression with shear
stiffness series

The stiffness value of the member plate could be successfully compression and shear stiffness in parallel seems to be in rea-

Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022

accounted for by the compression stiffness model without the sonable agreement with FEA prediction and hence, is the most
need for shear stiffness contribution from the member plate. This appropriate one to use.
makes logical sense because for the shear stiffness to play a role, • When no external surrounding constraints are applied to the
there should be a constraint parallel to the direction in which bolted members and they are free to deform under the action
force is applied. Without a shear constraint, the plate experiences of preload force, Model 1 which considers only the compres-
no resistance other than that governed by the compression of the sion stiffness seems to be in reasonable agreement with FEA
member. prediction and hence, is the most appropriate one to use.
• The disagreement between the analytical models and FEA
Relationship between member shear stiffness and external
curves increases significantly for H/d ratios less than one. It
constraints
can be seen that as H/d ratios < 1, Model 1 no longer corre-
a. Analytical and FEA models
sponds to the FEA results. The member compression stiffness
For cases where external constraints are added to the member formulas predict overly large values compared to FE. Since
plates clamped by a bolted connection, a simple bolted connection Models 2 and 3 also use the same member compression stiff-
spring model, which just takes compression stiffness into account, ness formulas, the effect is reflected in those models as well.
does not appear to be sufficient. An external fixed constraint on As the H/d ratio decreases, the member plate essentially tran-
one of the bolted members contributes a shear stiffness component sitions from a “thick” plate to a “thin” plate. It seems that the
to the bolted connection. To understand this better, three different member compression stiffness formulas break down in case of
spring models for Maruyama’s member stiffness experiment were plates that are thin relative to the bolt diameter. It is interesting
hypothesized as shown in Fig. 6. Model 1 accounted for stiffness to note that the breakdown starts at around the golden ratio
due to compression of the members only. Model 2 accounted the of 1.6.
both member compression and shear stiffness connected in parallel • Predictions from Model 3 (with shear and compressive stiff-
while Model 3 considered both the member shear and compression ness in series) do not correspond to FEA results.
stiffness to be connected in series. The results from analytical • The overall axial member stiffness tends to decrease initially as
models were compared to two FEA cases as represented in Fig. 7. the H/d ratio increases but then saturates after a point.
In the first case, no constraints were placed on the nodes surround-
ing the bolt head area while in the second case, constraints were d. Conclusions
present on the surrounding nodes. The stiffnesses for all these
The role of shear stiffness in the spring model of a bolted connec-
five cases were plotted as a function of H/d, where H was the thick-
tion depends on the state of external constraints imposed on the
ness of the plate and d was the bore diameter. The results were
bolted members. If the material surrounding a particular bolted
shown in Fig. 8.
member is fixed, there will be a shear stiffness component added
b. Calculation of member shear stiffness to the joint spring model due to that constraint.
For the case of the bolted bearing support structure (i.e., a pillow
The analysis to determine the member compression stiffness was
block), since fixed constraints are applied on the material surround-
outlined via Eqs. (1)–(10). Therefore, this section outlines the equa-
ing the bolted connection, a shear stiffness contribution from the
tions to calculate member shear stiffness. Slocum [13] derives a
bed can be expected. The overall member stiffness for this case
formula for calculating the shear stiffness of a bolted member
can be therefore calculated via Eq. (12).
given by Eq. (11).
1
loge 2 Kmember =
−1 (12)
Cbed|shear = (11) 1 1
2πtG Cflange|comp + +
Cbed|comp Cbed|shear
where, G is the shear stiffness of the flange member, and t is the
thickness. The thickness “t” in Eq. (11) is really the effective thick-
ness of the bolted member which is experiencing shear strain. From b. Bolt stiffness calculation:
Eq. (11), one might argue that as the thickness of the bolted member
As per Shigley [11], the bolt stiffness is given by Eq. (13). Con-
keeps increasing, the shear stiffness should keep increasing.
servatively, the area of the threaded portion (At) of the bolt is based
However, after a point, the shear stiffness of the bolted member sat-
on the root diameter. Table 8-2 from Shigley [11] gives a higher
urates because the additional thickness outside the region of influ-
order estimate of At based on the average of the pitch and root diam-
ence does not contribute to shear stiffness. A shear strain analysis
eters. For 1/4-20 UNC bolt, At was 0.0318 in.2 or 20.52 mm2.
plot for the member stiffness experiment conducted by Maruyama
[16] shows that the region of influence of the preload force Ad At E
extends only up to a limited number of elemental layers. Kbolt = (13)
Ad lt + At ld
c. Results and observations

• In cases where external surrounding constraints are applied on πd2

Ad = (14)
the bolted members, Model 2 which considers the member 4

ASME Open Journal of Engineering 2022, Vol. 1 / 011027-5

 Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022
Fig. 7 Constrained and unconstrained FEA models used for shear stiffness investigation: (a) FEA model with no constraints
on surrounding nodes and (b) FEA model with fixed constraints on surrounding nodes

Fig. 8 Stiffness as a function of cone parameters

011027-6 / Vol. 1, 2022 Transactions of the ASME

c. Total axial stiffness calculation: The pitch stiffness is thus realized as follows:
The bolt provides preload to the joint and its stiffness will be in Klinear δW
parallel to the stiffness of the material it compresses. Therefore, the M 6 Klinear W 2
Kpitch = = = (22)
axial stiffness of the joint is given by Eq. (15). θ δ 12
Kaxialjoint = Kmember + Kbolt (15) W/2
Here, the assumed pressure profile is symmetric and acts across
2.4.2 Pitch Stiffness Calculation. Next, a relation between the the whole width. Also, the Klinear, in this case, is equal to Kaxialjoint.
axial joint stiffness and the pitch stiffness needs to be established.
The pressure distribution profile at the joint interface is shown in b. Transformation of pitch stiffness to linear stiffness along
Fig. 9. A model presented by Slocum [13] for roller bearings is the pillow block axis:
used here for the bolted joint and it assumes a similar pressure dis- Equation (22) suggests that for pitch stiffness of the bolted con-
tribution profile to develop a differential stiffness expression and nection, the width of the member on which the moment is being
derive an equation relating the axial stiffness of the bolted joint to applied and the axial stiffness of the bolted connection play a

Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022

the pitch moment stiffness. major role.
a. Derivation of relation between axial stiffness and pitch Assuming, H to be the total height of the pillow block and Hc to
stiffness: be the height of the pillow block axis from the fixed base, the equiv-
alent stiffness of the structure in the direction of the force can be cal-
Consider an infinitesimal element (dx) of the pressure profile culated by Eq. (23).
shown in Fig. 9. Since the pressure profile is linear, the pressure
value of this element located at a distance x from the center of the Kpitch Klinear W 2
width of the pillow block can be found by multiplying the slope Kpitch(along the direction of force) = = (23)
Hc H 12Hc H
of the line with the distance from the center of the pillow block
width. The resulting force that this infinitesimal element exerts on
the base is given by Eq. (16). 2.5 Total Stiffness Along the Pillow Block Axis. As shown in
Fig. 2, all the four calculated stiffness values are in series. There-
Pext fore, the total axial stiffness of the pillow block can be written as
dF = Lxdx (16)
W/2 follows:
The moment equation can be written as per Eq. (17). 1 1 1 1
= + +
Pext 2 Ktotal|axial Khousing|bending Khousing|shear Khousing|torsional
dM = 2∗ Lx dx (17) 1
W/2 + (24)
Kpitch(along the direction of force)
Integrating to calculate the total moment,
 For the derivation of the pillow block body stiffness (including
4Pext L W/2 2 housing bending, shear and torsional stiffnesses), please see
M= x dx (18) Supplemental Material on the ASME Digital Collection.
W 0

W/2
4Pext L x3 Pext LW 2
M= = (19) 3 Experimental Setup
W 3 0 6
Figure 11 shows an overview of the experimental setup. A pillow
Assuming that a pressure Pext acting on a contact area of width W block made of plain carbon steel was bolted onto an aluminum base.
and length L causes a deflection δ, the linear stiffness (Klinear) can be The preload on the bolt was set to the desired level. For a maximum
given by force of 360 N along the pillow block axis, the required bolt preload
was calculated to be 1350 N, which corresponded to a torque
Pext LW requirement of about 5.68 Nm. Detailed calculations for critical
Klinear = (20)
δ loads to prevent loss of preload and gross slip can be seen in the
As shown in Fig. 10, using the small-angle approximation, δ = Supplemental Material on the ASME Digital Collection. An
θ*W/2. Substituting Eq. (20) in Eq. (19), aramid rope is used to pull so that no moment is applied on the
pillow block. A linear variable differential transformer (LVDT)
Klinear δW sensor with 100 nm least count is used to measure the deflection
M= (21) of the tip of the pillow block.
6

Fig. 9 Derivation of relation between linear and pitch stiffnesses

ASME Open Journal of Engineering 2022, Vol. 1 / 011027-7

 Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022
Fig. 10 Relation between angular and linear deflections

A digital spring scale was used to measure the applied force. The with reasonable accuracy close to the solid bolt model. The spider
pillow block assembly was held in place on the granite block using bolt model represents the bolt head/nut and shank with line ele-
two C clamps. The LVDT sensor was mounted on the base plate ments. The bolt shank can be modeled using a CBAR line
itself via a magnetic base indicator holder and the magnetic base element, which is essentially a simplified circular beam element
was attached to the test assembly via an intermediate plate of low- with stiffness governed via the cross section area and material prop-
carbon steel. This was done to avoid parasitic errors associated with erties defined by the user. The line elements representing the bolt
the deflection of the base plate itself with respect to the granite head/nut are modeled using RBE3 elements. RBE3 are rigid body
surface. The deflection of the pillow block was to be measured at elements generally used in finite element analysis for load transfer.
its top in the direction of application of force. Both CBAR and RBE3 elements are standard elements in the FEA
library. The line elements representing the bolt head/nut look like a
spider web connected to mesh structure of the flange members,
4 Finite Element Analysis Studies hence named as spider bolt model. In this FEA study, the bolted
FEA studies were conducted using NX 11 advanced simulation connections are represented using the spider bolt model.
software which is based on NASTRAN. Throughout the literature,
many different FEA models have been used to simulate bolted con-
nections. Some of these models are no-bolt, coupled bolt, spider 4.1 Finite Element Analysis Model Setup
bolt, and solid bolt. While the solid bolt model appears to be the 4.1.1 Meshing and Finite Element Method Model Creation.
best simulation approach for accuracy, it requires extra computation Various element types were used to model the components of the
effort and use of contact elements at interfaces [17]. The spider bolt bolted pillow block system. CTETRA(10), a quadratic 3D solid
model, on the other hand, is much more computationally efficient element, was used to model the base and the pillow blocks.

Fig. 11 Experimental setup

011027-8 / Vol. 1, 2022 Transactions of the ASME

 Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022
Fig. 12 Spider bolt model: Mesh construction details

CBAR element was used to model the bolt shank whereas RBE3 4.1.2 Application of Loads and Constraints. Figure 13 shows
elements were used for the spider connection. An element quality an overview of the loads and constraints applied in a general
check of the mesh indicated no failed elements. bolted connection model. The Y-direction force is applied along
Figure 12 shows the spider mesh used to create the bolted con- the inner circumference of the pillow block bore and bolt preload
nection. In a spider bolt model, the head and the nut (or threaded force is applied on each bolt. The bolt preload was represented in
hole) are modeled as rigid body elements connected to the the FEA model using the bolt preload feature in NX Advanced
clamped members. The bolt shank is represented as a bar element Simulation. This feature helps engineers apply bolt preload
(with the same cross section area). The spider mesh in the base forces. To achieve bolt-pretension, the FEA software essentially
plate tapped hole extends upto a depth equal to the thread engage- cuts the meshed body into two and uses pretension elements to
ment of the bolt (equal to approximately 1.3 bolt diameters). The write a constraint equation that relates the displacement of the cut
other end of the spider mesh connects to surface nodes on the boundaries accordingly [18]. In the FEA model and the experiment,
pillow block at the bolt head interface. The radial extent to which the base plate was fixed at the side tabs as indicated in Fig. 13. There
these connections are to be made is defined based on the diameter are four such regions in total (two on the top and two on the
of the bolt head. bottom). The interface area of the tabs was same as that of the

Fig. 13 Spider bolt model: Loads and constraints

ASME Open Journal of Engineering 2022, Vol. 1 / 011027-9

 Table 2 Final results summary (comparison of analytical, FEA, and experimental results for all
three pillow blocks)

Stiffness of the pillow block along the direction of

applied transverse load Error of the analytical model

Pillow blocks Analytical FEA Experimental w.r.t FEA w.r.t Experimental

2-bolt steel 70.8 N/um 75 N/um 60 N/um −5.6% 18%

4-bolt steel 133.6 N/um 126.6 N/um 113 N/um 5.5% 18.2%
2-bolt PLA 5.6 N/um 7.4 N/um 6.6 N/um −24.3% −15.1%

clamps used in the experimental testing. A surface-to-surface 6.1 Machine Tool Linear Axis

Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022

contact simulation object was added to the model. This simulation 6.1.1 Introduction. Machine tool axes are subject to very high
object creates contact elements at the pillow block-base plate inter- forces and thus require high stiffness actuators which are most often
face. Contact modeling requires the definition of a pair of surfaces ball screws. Guidelines and formulas are available to predict the
called the source and the target. If the mesh size for both the sur- stiffness of the screw shaft, nut, and support bearings, but analytical
faces is different, it is recommended that the surface with smaller models are needed to estimate the bearing housing stiffness which is
mesh size, be the source. This is because contact check is always usually bolted on a stiff mounting base. Often there is an offset
performed from the source side towards the target side. Normal pro- between the line of action of ball screw thrust force and the interface
jections are extended from the nodes on the source surface to the between the bearing block and the base to which it is bolted. This
target surface. Hence, if the source surface has smaller mesh size, results in a moment load being applied on the bolted bearing
the contact check can be performed with greater accuracy. No slip- block which in turn contributes to a load induced axial positioning
ping was allowed in the model. Preload for each of the bolts was set error in the system. Contact stiffnesses at the various joint interfaces
to 1350 N. of a machine tool also play a vital role in determining its dynamic
performance [20].

5 Verification of the Analytical Model Via Finite 6.1.2 Selection of Stiffness Matched Bearing Blocks. A spread-
sheet, presented in Supplemental Material on the ASME Digital
Element Analysis and Experiments
Collection, has been created to aid the designer in selecting the
A summary of the analytical, FEA, and experimental results for appropriate bearing block housing based on the stiffness values of
all three pillow block cases is shown in Table 2. The analytical other ball screw drive system components. A calculation flowchart
model is in reasonable agreement with the FEA predictions and is shown in Fig. 14. First, the operating parameters of the ball screw
experimental data. For detailed results and discussions on each drive system like the move profile, travel distance, carriage mass,
pillow block case, please see Supplemental Material on the process force, etc., are entered. An initial “guesstimate” is made
ASME Digital Collection. for the screw shaft diameter, length, and material which determines
the screw shaft stiffness. The next step is to then to initially specify
the screw lead, nut, and support bearings. Generally, the screw lead
is chosen such that the inertia ratio is close to one for optimal power
6 Case Studies on Application of the Analytical Model
transmission. The spreadsheet then calculates the minimum
To demonstrate the use of the analytical model for moment stiff- required bearing block stiffness.
ness of bolted connections, two case studies are presented that focus A plot of total system stiffness and the bearing block stiffness is
on ball screw bearing support structures (i.e., pillow blocks): automatically updated in the spreadsheet according to the parame-
a. Machine tool linear axis—This case study highlights the ters entered. After finding the minimum recommended stiffness
selection of an appropriate bearing support block based on values for the bearing blocks, the designer then enters information
stiffness matching with other components in ball screw regarding the bearing block geometry, mounting base, and the bolts
drive (screw, nut, and bearings) to meet the desired perfor- used to secure the bearing block. The spreadsheet calculates the
mance requirements for a machine tool linear axis. actual bearing block stiffness based and tries to find the closest
b. High-speed 3D printer—This case study highlights the selec- matching bearing block from a pre-existing catalog of various
tion of a ball screw support for a high-speed 3D printer. bearing blocks (e.g., from McMaster-Carr [21]). The designers
also have the capability to expand this catalog according to their
The stiffness contribution from ball screw drive components is needs.
typically assumed from catalog values for nut stiffness, if even Designers can also refer to Fig. 15 which shows a 3D design plot
available, or ignored in conventional models, and just shaft stiffness of the recommended bearing block stiffness as a function of the
is considered as the dominant compliance. Various research studies length and diameter of the selected ball screw shaft. Using this
have proposed better alternative stiffness models by closely plot, the designer can choose the best combination of the screw
accounting for such contributions from components/aspects shaft, screw diameter, and bearing block stiffness while meeting a
ignored in the conventional models. For example, Okwudire [19] particular system stiffness requirement. This plot was generated
proposed an improved screw-nut interface model by considering for reference only by using reasonable estimates for the stiffness
the effect of elastic deformation of the screw portion inside the of the nut and the support bearings. A design engineer could
nut. This model enabled a better prediction of the natural frequen- adjust stiffnesses accordingly for the intended application.
cies of the ball screw drive system. Similarly, bearing support
blocks are an essential component in the structural loop whose stiff- 6.1.3 Representative Bearing Blocks. Representative bearing
ness contribution has typically been ignored in conventional models blocks (McMaster-Carr catalog [21]) were considered where the
or at best is assumed to be equal to the nut stiffness. Since bearing two parameters most significantly affecting the pitch stiffness of
blocks are generally bolted onto the machine frame, the analytical bearing blocks are the diameter of bolts used to secure the
model for moment stiffness of bolted connections proposed in bearing block to the base and the ratio of the width of the block
this paper can be applied to a ball screw bearing support block to (W ) to its center height (Hc). Figure 16 plots the bearing block stiff-
account for its effect on the drive system stiffness and dynamics. ness as a function of the bolt diameter and the bearing block

011027-10 / Vol. 1, 2022 Transactions of the ASME

 Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022
Fig. 14 Ball screw analysis spreadsheet—calculation flowchart

geometry parameters (W/Hc ratio). The light coloured circles repre- different groups of pillow blocks. Table 3 details the cost and stiff-
sent 4-bolt bearing blocks and the blue circles represent 2-bolt ness ranges for these four groups. Group 1 represents the highest
bearing blocks. The size of the circles corresponds to the cost of stiffness pillow blocks which use big diameter bolts and have a
the bearing blocks. The numerical values next to the circles repre- high W/Hc ratio. Groups 2 and 3 are medium stiffness pillow
sent the stiffness of the bearing blocks in N/um. blocks that either have a high W/Hc ratio or a use bigger diameter
Is it better to have a 4-bolt bearing block with smaller bolt dia- bolt. Group 4 is low stiffness pillow blocks, which has both low
meter or a 2-bolt bearing block with bigger bolt diameter? Or W/Hc ratio and smaller bolt diameters. Although Group 4 bearing
what is the cost to stiffness trade-off between various types of blocks have low stiffness, they offer a compact footprint for
bearing blocks? Fig. 16 seeks to answer such questions and serve systems which have volume limitations.
as a reference tool for designers to balance the cost and stiffness
trade-off among various bearing block options. Although the stiff-
ness values are calculated assuming that the bearing blocks will 6.2 High-Speed 3D Printer. The objective of this case study
be mounted on an aluminum base using steel bolts, the user can is to describe how to select an appropriate ball screw support for
update these numbers in the spreadsheet to generate a trade-off a high-speed 3D printer. 3D printers are generally not subjected
plot representing their application more accurately. One of the to large external loads, but this can lead to a false sense of security,
key observations from this graph is that there are generally four and some end up being undersized for the dynamic performance

Fig. 15 Required bearing block stiffness as a function of ball screw shaft length and diameter

ASME Open Journal of Engineering 2022, Vol. 1 / 011027-11

 Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022
Fig. 16 3D scatter plot of bearing block stiffness as a function of bolt diameter, W/Hc ratio, and cost

Table 3 Cost versus stiffness comparison of various bearing block types (mounting base material: aluminum)

Group # W/Hc ratio Bolt diameter (mm) Stiffness range (N/um) Cost range (USD)

1 High (1.2–1.4) High (15–25) High (∼2000 or more) ∼1400 or greater

2 Low (1–1.2) High (15–25) Medium (600–750) ∼600–700
3 High (1.2–1.4) Low (6–15) Medium (600–900) ∼500–650
4 Low (1–1.2) Low (6–15) Low (300–400) ∼400–700

required. Ball screw assembly dynamics can significantly limit the stiffness, may be limited thereby placing greater stiffness require-
throughput. ments on the bearing support block. In addition, high lead/diameter
While a smaller diameter ball screw reduces system inertia ratio screws are often selected which can lead to effective axial stiff-
leading to faster response times, it also reduces the screw shaft stiff- ness decrease due to torsional windup of the shaft as derived by
ness leading to lower natural frequencies. Axial vibrations are just Slocum [13].
one of the many modes of vibration that the ball screw shaft is sus- The throughput of a 3D printer can be assumed to scale along the
ceptible to. High lead/diameter ratio screws also suffer from tor- lines of the maximum achievable angular velocity of the ball screw
sional windup that must be considered as part of the effective shaft. Similarly, one of the contributors to the cost of a 3D printer is
axial stiffness of the screw (Slocum [13]) Due to the coupling intro- the servomotor/amplifier package which generally scales with the
duced by the nut, the application of torque on the ball screw can also power rating of the motor.
cause unwanted lateral vibrations of the ball screw, thereby affect-
ing the system’s positioning accuracy. Okwudire [22], in fact,
showed analytically that while the torque induced lateral vibrations
cannot be completely eliminated, but they can be reduced by wisely 6.2.1 Throughput Dependency on the Bearing Block Stiffness.
choosing the entry/exit angles of the balls in the ball screw drive Suppose a designer’s objective is to maximize the throughput of
mechanism. Slocum [13] discusses many of the factors affecting the 3D printer while keeping the same motor power and system
well-known optimal transmission ratio formulas for two categories natural frequency. Since the ball screw shaft and the bearing
of systems: ones which encounter large external forces during oper- block act as springs connected in series with one another, the
ation and others for which the encountered external forces are neg- designer can try to minimize the ball screw shaft diameter and max-
ligible. Since 3D printers fall in the latter category, the simple imize the bearing block stiffness to maintain the same system
inertia dominated formula may be used to determine the optimal natural frequency. Reducing the diameter of the ball screw shaft
transmission ratio (n) as shown in Eq. (25). reduces the rotary inertia of the system which in turn leads to a
 higher achievable angular acceleration. Figure 17(a) shows how
Jload the maximum achievable angular velocity increases as a function
Optimal|Transmission|Ratio|(n) = (25)
Jmotor of the bearing block stiffness. Quite predictably, the benefit of
using a stiffer bearing block wanes after a certain point. While stif-
For a direct drive connection between the motor and shaft (trans- fening the bearing block helps limit impact to the natural frequency
mission ratio = 1), optimal power transmission is achieved when the for the axial vibration mode, the natural frequencies for other vibra-
inertia of the motor is equal the inertia of the driven load, but when tion modes (e.g., torsional and lateral modes) may still be impacted
satisfying the inertia ratio, the shaft diameter, and hence axial significantly with reduction in shaft diameter.

011027-12 / Vol. 1, 2022 Transactions of the ASME

 Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022
Fig. 17 Throughput and cost dependency on bearing block stiffness: (a) maximum angular velocity as a function of bearing
block stiffness and (b) minimum required motor power as a function of bearing block stiffness

6.2.2 Cost Dependency on the Bearing Block Stiffness. Using a designer uses, there is a minimum limit to the required motor
stiffer bearing block with a smaller diameter ball screw shaft may power based on the required move profile and other operating
also be a good strategy to reduce the cost of a 3D printer without parameters.
impacting its throughput. If the same strategy of minimizing the
rotary inertia of the system while maintaining the system natural fre-
quency is followed, the designer will need a motor with lower
power rating to achieve the same angular velocity of the shaft.
Reducing motor size or power reduces the cost of the servomotor 7 Conclusions
and amplifier package accordingly [13]. Figure 17(b) shows the An analytical model for predicting the pitch stiffness of bolted
required motor power as a function of the bearing block stiffness. connections has been presented. This model is verified by finite
It can be seen that the minimum required motor power falls drasti- element analysis and experimental testing. Three different pillow
cally at first as the bearing block stiffness is increased but then the blocks were used as test cases for this verification. For all three
curve saturates very quickly. Based on Fig. 17(b), an optimal choice cases, the load deflection curves for analytical, experimental, and
for a bearing support block in this case will be one with a stiffness of FEA results are in reasonable agreement. Observations from these
at least 100 N/um. Irrespective of the type of bearing block the studies include as follows:

ASME Open Journal of Engineering 2022, Vol. 1 / 011027-13

 • Beware subtle sources of compliance in the experimental consent is not applicable. This article does not include any research
setup. in which animal participants were involved.
• Micro slip may occur at the pillow block-base plate interface.
To make sure that the measurements can be attributed to elastic
system stiffness, one should check that there is no residual
deflection once the system is unloaded. Data Availability Statement
• It is important to choose test cases where the dominant source
of compliance is the focus of the study. It is very difficult to The datasets generated and supporting the findings of this article
isolate individual sources of compliance during experimental are obtainable from the corresponding author upon reasonable
testing. request.
• For a multi-bolt pattern, it is important to follow the recom-
mended tightening sequence (according to whether the
pattern is circular or non-circular). At least two tightening
cycles should be used across the bolt pattern to minimize var- References

Downloaded from http://asmedigitalcollection.asme.org/openengineering/article-pdf/doi/10.1115/1.4054474/6898742/aoje_1_011027.pdf by guest on 01 November 2022

iation in preload due to bolt cross-talk. [1] Zhang, O., and Poirier, J. A., 2004, “New Analytical Model of Bolted Joints,”
• It is important to be aware of geometrical limitations of the ASME J. Mech. Des., 126(4), pp. 721–728.
object of interest (e.g., pillow block in this case). The strain [2] Nassar, S. A., and Abboud, A., 2009, “An Improved Stiffness Model for Bolted
cone field saturates prematurely if the bolts are too close to Joints,” ASME J. Mech. Des., 131(12), p. 121001.
[3] Wileman, J., Choudhury, M., and Green, I., 1991, “Computation of Member
the edges of the pillow block. In such cases, one should try Stiffness in Bolted Connections,” ASME J. Mech. Des., 113(4), pp. 432–437.
estimating the stiffness of the flange members as per their geo- [4] Lehnhoff, T. F., Ko, K. I., and McKay, M. L., 1994, “Member Stiffness and
metrical limitations. Contact Pressure Distribution of Bolted Joints,” ASME J. Mech. Des., 116(2),
• In an FEA study, close attention should be paid to the finite pp. 550–557.
[5] Sethuraman, R., and Sasi Kumar, T., 2009, “Finite Element Based Member
element mesh. Refining the mesh at the important interfaces Stiffness Evaluation of Axisymmetric Bolted Joints,” ASME J. Mech. Des.,
and verifying mesh convergence are important steps to get 131(1), p. 011012.
accurate and trust-worthy results. [6] Ito, Y., and Masuko, M., 1971, “Study on the Horizontal Bending Stiffness of
• Close attention should be paid to the external constraints Bolted Joint,” Bull. JSME, 14(74), pp. 876–889.
[7] Gould, H. H., and Mikic, B. B., 1972, “Areas of Contact and Pressure Distribution
applied on the bolted joint system and how it affects the in Bolted Joints,” ASME J. Eng. Ind., 94(3), pp. 864–870.
overall joint stiffness model. Section 2.4.1 detailed how the [8] Oskouei, R. H., Keikhosravy, M., and Soutis, C., 2009, “Estimating Clamping
contribution of the member shear stiffness to the overall Pressure Distribution and Stiffness in Aircraft Bolted Joints by Finite-Element
bolted joint stiffness depends on the state of the external con- Analysis,” J. Aerosp. Eng., 223(7), pp. 863–871.
[9] Naruse, T., and Shibutani, Y., 2012, “Nonlinear Bending Stiffness of Plates
straints imposed on the members. Clamped by Bolted Joints Under Bending Moment,” J. Solid Mech. Mater.
• The application of the presented analytical model for selection Eng., 6(7), pp. 832–843.
of ball screw bearing support blocks has been illustrated via [10] Precision Machine Components, NSK Ltd., Tokyo, Japan, 2017, pp. B17–B18.
two case studies: one is based on a machine tool linear axis [11] Budynas, R., and Nisbett, J., Shigley’s Mechanical Engineering Design, 9th ed.,
McGraw Hill.
and another is based on a high-speed 3D printer. The capability [12] Williams, J. G., Anley, R. E., Nash, D. H., and Gray, T. G. F., 2009, “Analysis of
of predicting the stiffness of bearing support blocks may help Externally Loaded Bolted Joints: Analytical, Computational and Experimental
designers to optimize a ball screw system as per their system Study,” Int. J. Press. Vessel. Pip., 86(7), pp. 420–426.
requirements, whether it be maximizing system throughput, [13] Slocum, A. H., 1995, Precision Machine Design, Society of Manufacturing
Engineers, Dearborn, MI, pp. 365–393.
reducing cost, improving the system’s natural frequency, or [14] Du, F., Li, B., Zhang, J., Zhu, Q. M., and Hong, J., 2015, “Ultrasonic
reducing load induced positioning errors. Measurement of Contact Stiffness and Pressure Distribution on Spindle–Holder
Taper Interfaces,” Int. J. Mach. Tools Manuf., 97, pp. 18–28.
The analytical models developed in this paper can be readily [15] Cao, J., and Zhang, Z., 2019, “Finite Element Analysis and Mathematical
incorporated into a design spreadsheet, as part of a product’s Characterization of Contact Pressure Distribution in Bolted Joints,” J. Mech.
design history file, for rapidly considering “what-if” scenarios. Sci. Technol., 33(1), pp. 1–11.
[16] Maruyama, K., Yoshimoto, I., and Nakano, Y., December 1975, “On Spring
Later, once a workable design has been developed, it can be fine- Constant of Connected Parts in Bolted Joints,” Bull. JSME, 18(126), pp. 1472–
tuned with a more elaborate finite element model. 1480.
[17] Montgomery, J., 2002, Methods for Modelling Bolts in Bolted Joints, Seimens
Westinghouse Power Corporation, Orlando, FL, Technical Report, p. 8.
[18] Oatis, D., 2007, “Analyzing Bolt Pretension in the ANSYS Workbench
Acknowledgment Platform,” ANSYS Adv, 1, pp. 28–29.
The authors would like to thank John Massucci, Geoffrey Wil- [19] Okwudire, C. E., 2011, “Improved Screw–Nut Interface Model
for High-Performance Ball Screw Drives,” ASME J. Mech. Des., 133(4),
liams, and Phil LaMastra from ASML US LP for providing p. 041009.
access to the computational resources and FEA software (NX [20] Chang, Y., Ding, J., He, Z., Shehzad, A., Ding, Y., Lu, H., Zhuang, H., Chen, P.,
Advanced Simulation 11) used in this study. The authors would Zhang, Y., Zhang, X., and Chen, Y., “Effect of Joint Interfacial Contact Stiffness
also like to thank John Fortunato, Jim Chester, Andy Hamilton, on Structural Dynamics of Ultra-Precision Machine Tool,” Int. J. Mach. Tools
Manuf., 158, p. 103609.
and Dragos Pariza from ASML US LP for their valuable advice [21] “End Supports for Lead Screws and Ball Screws,” McMaster-Carr, https://www.
and suggestions. This research did not receive any specific grant mcmaster.com/bearing-blocks/end-supports-for-lead-screws-and-ball-screws-4/,
from funding agencies in the public, commercial, or not-for-profit Accessed January 5, 2021.
sectors. [22] Okwudire, C. E., 2012, “Reduction of Torque-Induced Bending Vibrations in Ball
Screw-Driven Machines Via Optimal Design of the Nut,” ASME J. Mech. Des.,
134(11), p. 111008.
[23] Juvinall, R., and Marshek, K., Fundamentals of Machine Component Design, 5th
Conflict of Interest ed., John Wiley and Sons, Inc., Hoboken, NJ, p. 419.
[24] Ghosh, S., 2011, Typical Coefficient of Friction Values for Common Materials,
There are no conflicts of interest. This article does not include MechGuru, https://mechguru.com/machine-design/typical-coefficient-of-friction-
research in which human participants were involved. Informed values-for-common-materials/, Accessed November 15, 2018.

011027-14 / Vol. 1, 2022 Transactions of the ASME