You are on page 1of 7

IADC/SPE 87190

Application of Increased Make-up Torque on API Rotary-Shouldered Connections:
Goodman Diagram vs. Strain Life Model
John P. McCarthy, SPE, S. DeWayne Everage, SPE, Kang Lee, PhD, SPE, T H Hill Associates, Inc.

Copyright 2004, IADC/SPE Drilling Conference
This paper was prepared for presentation at the IADC/SPE Drilling Conference held in Dallas,
Texas, U.S.A., 2–4 March 2004.
This paper was selected for presentation by an IADC/SPE Program Committee following
review of information contained in a proposal submitted by the author(s). Contents of the
paper, as presented, have not been reviewed by the International Association of Drilling
Contractors or Society of Petroleum Engineers and are subject to correction by the author(s).
The material, as presented, does not necessarily reflect any position of the International
Association of Drilling Contractors or Society of Petroleum Engineers, their officers, or
members. Papers presented at IADC/SPE meetings are subject to publication review by
Editorial Committees of the International Association of Drilling Contractors and Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the International Association of Drilling
Contractors and Society of Petroleum Engineers is prohibited. Permission to reproduce in print
is restricted to a proposal of not more than 300 words; illustrations may not be copied. The
proposal must contain conspicuous acknowledgment of where and by whom the paper was
presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A.,
fax 01-972-952-9435.

Abstract
This paper details the evaluation of rotary shouldered
connection (RSC) fatigue performance by comparing the
widely used Modified Goodman Diagram (MGD) to a more
advanced and accurate elastic-plastic Strain Life Model
(SLM). The technical approach defined in this paper
challenges the application of the MGD for made-up API rotary
shouldered connections. Likewise, the approach challenges
the general assumption that the make-up torque (MUT) values
defined in API RP7G are set in stone. These make-up torque
values do, however, have a consistent basis. Understanding
this basis, the state of stress in the connection and how a rotary
shouldered connection functions, allows us to use the
mechanics of the connection to our advantage. Such an
assessment recently led to a justification for increasing the
make-up torque on an API NC-50 connection for a major
operator while drilling a challenging directional well in the
Gulf of Mexico. This example is reviewed in light of both the
MGD and the SLM. A concerted effort to understand and
evaluate the connection performance allowed application of
increased MUT for reaching TD without having to lay down
drill pipe and incur the cost associated with switching to a
high-torque rental string. Any operator or service provider
can utilize the evaluation techniques presented in this paper to
assess the impact a controlled increase in make-up torque may
have on a rotary shouldered connection.
Introduction
Current drilling prospects continue to require operators, both
large and small, to extend the reach, depth, and complexity of
the wellbore to achieve the target objectives. This often
translates into higher mechanical loads - both torque and
tension – acting on the drill string. As a result, a shift in the

market is reflected by the introduction of several new
technologies and enhanced-feature tools to meet the increased
demands [1] [2]. High-torque drill strings, torque reducing subs,
and rotary steerable systems all aim at increasing the available
capacity under given geometric design constraints[3].
Practically speaking, one of the first load conditions that pose
a design challenge in deviated well trajectories is torque.
Certainly, there are situations where specialty drillstrings with
proprietary connections and rotary steerable assemblies are
justified, and in some cases required to mitigate, among other
things, increased operating torque. However, there also exists
those opportunities when simply using the available torsional
capacity of the standard (and readily available) API rotary
shouldered connection is a technically founded and cost
effective option. However, some continue to use the Modified
Goodman Diagram (MGD) as a means to show that the fatigue
life of a RSC is reduced simply by increasing MUT beyond
that specified by API. This paper provides an accurate
explanation for interpreting the Modified Goodman Diagram,
a proposed explanation as to why the MGD does not apply to
RSC’s in the made up condition, and an introduction to a more
accurate fatigue life estimation tool - the Strain Life Model
(SLM). This model will also be used to show that MUT has
little effect on the fatigue life of rotary shouldered
connections, in most situations encountered in the industry.
Make-up Torque (MUT)
A basic understanding of an API RSC is required before it is
possible to justifiably extend beyond the sound basis of the
API’s recommendations regarding make-up torque.
Comprehensive discussions on the basis for tool joint make-up
torque and the resulting states of stress at critical locations in
the connection geometry have been explored in detail in both
References 4 and 5 as well as in API RP7G and Standard DS1™.
Here we will repeat the important points
for convenience.
Make-up torque for rotary shouldered connections serves
several purposes:
1.
2.
3.
4.
5.

Connect joints of drill pipe
Contain pressure (internal and external)
Maintain connection integrity during bending
Transmit torque to the bit
Maintain tensile integrity while pulling on the
pipe

2

IADC/SPE 87190

The API recommended make-up toque for a rotary
shouldered connection is the torque required to achieve a
desired stress level in the weaker of either the pin or box based
on the smaller of the cross-sectional areas calculated at 0.750″
from the sealing shoulder on the pin and 0.375″ from the
sealing shoulder of the box. Make-up torque (ft-lb) is
calculated using Equation 1 below:

T=

R f
SA ⎡ P

+ t + Rs f ⎥

12 ⎣ 2π cos Φ
⎦ ………………………... (1)

Where:
T
S
A
P
Rt
Rs
f
Ф
Tpr

= Make-up Torque (ft-lb)
= Desired stress level from make-up
= Cross-sectional area (in2)
= Lead of threads (in)
= Average mean radius of thread (in)
= Mean shoulder radius (in)
= Thread compound coefficient of friction
= ½ thread angle (Figure 9.2 or 9.3, API Spec. 7)
= Thread taper, on the diameter, (in/ft)

For standard API tool joints [120 ksi minimum yield
strength (MYS)], the “recommended” make-up torque is based
on 60% of the MYS – a target bulk stress of nominally 72 ksi.
The average stress is far lower than the actual stress at the
root of the last engaged thread (LET) on the pin (roughly 3/4"
from the shoulder). In fact, the root of the LET is stressed
beyond yield at or near the API recommended MUT. For
example, based on finite element analysis (FEA), a NC-38
connection, with 5" OD, and 2 11/16" ID will have a peak
stress of over 160 ksi at the root of the LET when 9,000 ft-lb
of make-up torque is applied. For additional information on
calculating actual make-up torque values using Equation 1
above, reference API RP-7G Appendix A.8 or Standard DS1™, 2nd edition, Appendix A.4.1. Further elaboration here
would simply be a repeat of these and other widely
referenced publications.
With the basic premise for how make-up torque is
calculated now defined, the next important concept to
investigate is the RSC Combined Load Curve (CLC). Details
related to this material are also well documented, so will not
be fully explained in the context of this paper. Further
reference can be found in Standard DS-1™, 2nd edition,
Appendix A.4.3, with detailed formulas provided.
In cases where more make-up torque is desired than is
“standard”, based on the 72 ksi target stress from API,
evaluating a CLC for the subject connection allows a user to
make informed decisions regarding how increased make-up
torque will effect the tensile and torsional capacities of the
connection. When a connection is torqued above the standard
72 ksi target stress level, several questions must
be considered:
1.
2.

Is the connection pin or box weak in torsion?
How does the additional tensile preload in the tool
joint affect the connection’s ability to carry tension?

3.
4.
5.

What is the tool joint’s remaining tensile capacity at
the applied MUT?
What is the bearing stress at the sealing shoulder?
Will the fatigue life of the connection be reduced due
to the increased MUT?

The answer to question five will be pursued in the remainder
of this paper.
Failures
In today’s oilfield, drill pipe fatigue failures almost always
occur in the tube, not the tool joint. Before API Numbered
Connections (NC) were introduced in 1968, the failure-prone
V-065 thread form was standard on many drill pipe connection
types (all IF connection types used this thread form). These
connection types, with flat bottom thread roots and 0.015"
radii at the corners were prone to fatigue. To compound the
problem, make-up torque was not applied consistently. As
noted in Reference 6, common oilfield practice in the 1950’s
for applying make-up torque was to wrap the tong line around
the cat-head one time for each inch of tool joint OD, which led
to under-energized connections. Torque gages were not used,
drilling torque was not considered, and many failures
occurred. The primary failure mechanism was fatigue in the
pin, caused by pin wobble from low MUT, coupled with the
stress concentration resulting from the V-065 thread form.
An API committee was formed in 1964 to address and
unify connection design as well as make-up torque. The result
was the NC connection. All NC connections employ the V038R thread form which has a single 0.038" thread root radius
and is much more fatigue resistant than the V-065. Since they
were introduced, NC connections have become the primary
connection used on drill pipe tool joints, and pin fatigue
failures have been virtually eliminated. Most fatigue failures
in joints of drill pipe now occur in the tube, near the internal
upset, or in the slip area. In fact, having investigated several
hundred drill pipe failures, T H Hill Associates, Inc. (THHA)
has yet to see a fatigue failure in a tool joint pin with the V038R thread form.
Even though fatigue failures of drill pipe tool joints are
very rare, the fatigue resistance of a connection must still be
evaluated for drill string design purposes. Reference 15
provides more detail in designing for and mitigating fatigue.
While the MGD is a valid method when plastic strain is not
present, the SLM is the better choice when evaluating tool
joints, for fatigue, in the made-up condition. As such, the rest
of this paper will show that the MGD does not apply to madeup RSC’s, because plastic strain is present at normal MUT
values, and that increasing MUT above API recommended
values typically has little effect on the fatigue life of
a connection.
Modified Goodman Diagram
The Modified Goodman Diagram (MGD) is a stress based,
analytical model used to evaluate the endurance limit of a
metal, based on mean and cyclic stresses. It uses an average
or bulk stress over the entire cross section and applies a stress
concentration factor when notches or other stress raisers are
present. It is accurate for small, elastic strains, or when
applied to shafts and other shapes with known stress

IADC/SPE 87190

3

concentration factors[7]. The MGD is limited to the elastic
region. If plastic strain exists, as is the case with RSC’s at
MUT values near and above those specified by the API, the
MGD should no longer be the method of choice for predicting
fatigue life.
Before a more accurate alternative method for predicting
fatigue life is introduced, it is useful to review the way the
MGD is commonly used. Reference 8 provides a detailed
explanation of how the MGD is constructed, and will not be
repeated here. The main point to remember from the MGD is
that any combination of mean stress and cyclic stress that falls
below the Modified Goodman line should yield infinite life.
This implies that as long as the combined stresses are below
the line, no fatigue damage is done to the item in question.
Figure 1 shows a MGD for a drill pipe tool joint.
The primary parameter of the MGD is the stated endurance
limit of 13,000 psi cyclic stress amplitude (at zero mean
stress) defined at 1,000,000 cycles. This value is considered
the maximum cyclic stress that can be applied, to an un-madeup pin, and still obtain an “infinite” fatigue life (defined in
Reference 6 arbitrarily as 1,000,000 cycles), and was obtained
through laboratory testing in the 1950’s[6]. The testing was
most likely performed on a connection with the V-065 thread
form, although it is not clear from Reference 6, which
connection type was used.
20
18

Cyclic Stress (ksi)

16
14

Finite Fatigue Life

12
10
8
6

Infinite Fatigue Life

4
2
0
0

20

40

60

80

100

120

140

Mean Stress (ksi)

range generally employed on a RSC) has little impact on the
number of cycles to failure.
A common interpretation of the MGD is that increasing
MUT above API recommended values decreases the fatigue
life of a RSC. This is not the case. A more accurate
interpretation of the MGD (assuming one can apply the MGD
to a RSC in the made-up condition) is that the maximum
cyclic stress allowable for a RSC to obtain “infinite” life is
decreased as MUT (mean stress) is increased. In other words,
simply by increasing MUT, one does not necessarily reduce
the fatigue life of a RSC. Using the MGD, along with the
mean stress, a new endurance limit can be computed,
according to the Modified Goodman line. If one asserts that
the MGD can be applied to a RSC in the made-up condition,
as long as the imposed cyclic stress is below the newly defined
endurance limit, no fatigue damage will occur.
A recent example where the above assertion was
considered is useful in comparing the differences that exist in
MGD interpretation. A major operator was drilling a well in
the Gulf of Mexico and inadvertently put dog-legs of about 3.5
deg/100 ft relatively close to surface (5,000' MD). Torque and
drag modeling, based on actual drilling data indicated that
when approaching TD of the well, surface drilling torque was
going to exceed the MUT of the NC-50 connection (6 5/8" OD
by 3 1/4" ID) of the rig’s 5", 19.5 lb/ft, S-135 drill pipe.
THHA reviewed the drill pipe in question, and determined that
increasing the MUT from the rig’s value of about 28 kft-lb to
36 kft-lb would allow the well to be drilled successfully to TD
without the operator incurring the added costs of laying down
pipe and renting a high-torque string (API MUT for the
connection in question is 31 kft-lb). It should be noted that for
various reasons, the drilling contractor was initially using a
lower MUT than specified by API.
Following the
recommendation to increase MUT, there was some fear of
reducing the fatigue life of the drill pipe connection. Initially,
the MGD was provided as proof that increasing MUT would
decrease fatigue life. A detailed load evaluation was made on
the drill string at the point of the maximum dog-leg. See
Table 1 for a list of all stresses (based on the detailed load
evaluation) that would be acting on the connection at 5,000'
MD (across the dog-leg), while drilling at TD.
Table 1. Comparison of Two MUT Values on a NC-50
Connection and the Effect on Fatigue Life
Based on the MGD.

Figure 1 - Modified Goodman Diagram for a Drill Pipe Tool Joint
Case

In the 1950’s, 1,000,000 cycles may have been
considered infinite life, but today a drill string will see
1,000,000 cycles in just over 100 hours, when rotated at 150
RPM, which can easily be accomplished in a week’s drilling.
A more recent paper considered two million cycles as
“infinite”[8]. Today, with high speed top-drives and 30+ kft
deep wells, drill pipe can easily see 10 million cycles between
inspections. Therefore, the authors believe that 10 million
cycles should be the absolute minimum criteria for “infinite”
life. As can be seen from the examples above, the choice of
number of cycles needed for “infinite” life is somewhat
arbitrary. The point that will be made here is that MUT (in the

Tensile
Stress
(ksi)1

Stress
due to
internal
pressure
(ksi)2

Bulk
stress
due to
MUT (ksi)

Sum of all
Stresses
(ksi), (Mean
Stress)

MGD Max
Cyclic
Stress for
Infinite Life
(ksi)

28 kft-lb
8.6
0.41
65.0
74.0
6.2
MUT
36 kft-lb
8.6
0.41
83.5
92.6
4.4
MUT
1. Tensile stress based on 200 klb hanging on tool joint at 5 kft MD.
2. Stress due to internal pressure based on 4,000 psi pump pressure.

The maximum cyclic stress imposed on the tool joint by
rotating it through a 3.5 deg/100 ft dog-leg was calculated to
be about 3.7 ksi. Since the calculated cyclic stress expected
while drilling was less than the maximum allowable predicted
by the MGD (4.4 ksi), the conclusion was that ‘no fatigue

4

IADC/SPE 87190

damage’ would result by increasing the MUT to 36 kft-lb from
28 kft-lb. The MUT was increased and the well was drilled to
TD without incident. Figure 2 shows the MGD for the two
MUT cases, with the mean stresses plotted against allowable
cyclic stresses. The above example uses the methodology and
formulas presented in Reference 8 to calculate the stresses
acting on the tool joint. The only difference is in calculating
the stress acting on the tool joint due to bending caused by the
dog-leg. We conservatively assumed the tool joint would
conform to the dog-leg. Reference 8 assumes the tube
conforms to the dog-leg and transfers the corresponding
bending moment to the tool joint. If the methodology used in
Reference 8 were used to calculate the cyclic stress acting on
the tool joint, it would have yielded only 0.6 ksi bending
stress. In reality, the stress acting on the tool joint would have
been somewhere between 0.6 ksi and 3.7 ksi. By using the
most conservative approach for estimating the cyclic bending
stress, the tool joint should not have accumulated any fatigue
damage. Reference 11 provides a detailed discussion of how
drill pipe tubes and tool joints react when rotated through
boreholes with dog-legs.

20
18

Cyclic Stress (ksi)

16
14
12

28 kft-lb MUT
10

36 kft-lb MUT

8
6
4
2
0
0

20

40

60

80

100

120

140

Mean Stress (ksi)
Figure 2 - Modified Goodman Diagram showing the relationship
between two MUT cases. The dots represent the maximum
“imposed” cyclic stress for the operation considered.

Strain Life Model
The stress based MGD ignores the plastic deformation that
occurs in the root of the last engaged thread during normal
make-up of all RSC’s. It assumes the stress induced in the pin
3/4" from the shoulder is evenly distributed throughout the
cross section. This is not true. The thread root is stressed well
beyond yield and the stress diminishes with distance towards
the ID of the pin. Figure 3 shows the stress distribution of a
NC-38 connection when made-up to 9,000 ft-lb MUT. From
Figure 3, the root of the LET is stressed well beyond yield
(170 ksi vs. 120 ksi), which means plastic strain exists in
the LET.

Plastic strain at root of LET
Figure 3 - Stress Distribution of a NC-38 Connection at 9 kft-lb
MUT, Showing Plastic Strain at Root of LET.

When the connection is made-up to near API MUT and
higher, the stress in the last engaged thread root is so high that
a linear, stress based approach to fatigue life can no longer be
used to predict an endurance limit or fatigue life. The plastic
deformation in the thread root must be acknowledged and
accounted for when attempting to predict fatigue life.
A strain based approach can account for the plastic
deformation that occurs in the thread root of the last engaged
thread of the pin at MUT values used on RSC’s. The Strain
Life Model (SLM) predicts fatigue life based on the cyclic
stress-strain state at the likely locations of fatigue crack
initiation, such as the roots of threads in a RSC. However, a
finite element model (FEA) is required to accurately calculate
and model the stress-strain state in the pin of a RSC in the
made-up condition.
The recent progress in computer and software technology
makes it possible to use non-linear FEA to determine the true
cyclic stress-strain state in areas of high stress concentration,
such as the root of the LET. The FEA non-linear model works
with the SLM to provide accurate fatigue life analysis.
The modeling process begins by importing a machine
drawing of the connection into a FEA program. A mesh is
then made that distributes elements and nodes throughout the
connection. A finer mesh is made around areas with large
stress gradients and other areas of interest, such as thread
roots. The amount of pin stress due to MUT and other loads
such as tension and pressure are applied next (the sum of
which determines the mean stress in the connection). Finally,
bending is applied to the connection. To determine the inputs
for the SLM, stress and strain are then calculated for each
element and node.
The Morrow Strain Life Model is then used to calculate
the number of cycles to failure based on the stress-strain
distribution calculated with the FEA model. Different values
of cyclic stress are entered into the SLM, and the number of
cycles to failure is recomputed. Performing this sequence
several times allows curves to be generated that show the

IADC/SPE 87190

5

Modeling Results
Five FEA models were constructed and evaluated for a NC-38
tool joint (5" OD and 2 11/16" ID). Table 2 shows the stressstrain state at the LET due only to MUT, without any
cyclic loading.
Table 2. FEA Models Created for Additional Testing
on a NC-38 Connection with 5" OD and 2
11/16" ID
MUT
(ft-lb)
0
1,000
2,000
9,000
13,000

Mean
Stress at
LET (psi)
0
6,200
13,000
60,000
89,300

Peak
Stress at
LET (psi)
0
21,200
46,950
166,700
170,800

Mean
Strain at
LET (in/in)
0
0.00019
0.00039
0.0019
0.0030

Peak Strain
at LET
(in/in)
0
0.00064
0.0014
0.0088
0.014

The above load cases were loaded into the SLM and
cyclic loading was applied over a wide range for each case.
Curves were generated for each load case as shown in Figure
4. Table 3 shows the endurance limit extrapolated from the
curves (cyclic stress that yields 10,000,000 cycles to failure) in
Figure 4, as well as the endurance limit predicted by the MGD
based on the mean stress at the LET for each MUT case.
Table 3. Results of the Strain Life Model Compared
to the Modified Goodman Diagram
MUT (ft-lb)
0
1,000
2,000
9,000
13,000

MGD Endurance
Limit (psi)
13,000
12,400
11,750
7,300
4,700

SLM Endurance Limit
(psi)
15,000
12,800
11,400
5,200
5,200

According to the MGD, the maximum cyclic stress for
infinite life should have decreased from one load case to the
next. The FEA modeling results show there was not much of
a decrease in fatigue life due to increasing MUT from 9,000
ft-lb to 13,000 ft-lb, even though the MGD suggests that the
maximum allowable cyclic stress to ensure infinite fatigue life
should have been reduced by more than 35%. The FEA
correlates fairly well to the MGD for the first three load cases
because there is no plastic deformation in the connection.
However, once the connection is made-up to near API MUT
and above, the MGD will no longer predict infinite fatigue
life accurately.
4 50 0 0
40000

Cyclic Stress Amplitude (psi)

number of cycles to failure based on cyclic stress. From the
curves, an endurance limit can be visualized. For those who
prefer more information, Reference 12 provides more detail on
how the FEA model is constructed and how the SLM
is implemented.
With the SLM theory introduced, it is useful to apply it
to our example of the NC-50 connection and increased MUT.
A dimensionally identical connection (6 5/8" OD by 3 1/4"
ID) was input into the FEA model, the same load conditions
were placed on the connection, and the stress-strain
distribution was calculated. The SLM predicted almost the
same number of cycles to failure for the 28 kft-lb and 36 kft-lb
cases (5.47 x107 and 5.38 x107 respectively). This supports
the theory that once plastic strain is present in the LET, the
fatigue life is only marginally affected as MUT is increased.
Since the calculated number of cycles to failure was more than
five times the minimum value of 1 x 107 cycles considered for
“infinite” life in this paper, the actual encountered cyclic stress
of 3,700 psi was below the “endurance” limit.
Additional modeling was performed to further support
the theory that increased MUT has little effect on fatigue life
once a connection is made-up enough to induce plastic strain
in the root of the LET on the pin. Previous research and
testing was performed on NC-38 connections[8], so our
modeling was based on the same connection.

3 50 0 0
30000
2 50 0 0
20000
150 0 0
10 0 0 0
50 0 0
0
1.E+0 3

1.E+0 4

1.E+0 5

1.E+0 6

1.E+0 7

1.E+0 8

1.E+0 9

1.E+10

C yc le s to F a ilure
9 kft -lb FEA-SLM
13 kft-lb FEA-SLM
2 kft -lb FEA-SLM

0 ft-lb FEA-SLM
1 kft-lb FEA-SLM

Figure 4 - FEA Strain Life Modeling Results (NC-38 Connection)

The experimentally determined endurance limit for a
RSC, in the un-made-up condition may very well be near
13,000 psi[6], and is likely higher for a RSC with the V-038R
thread form. The SLM approach and actual fatigue tests [8]
indicate that the endurance limit is closer to 15,000 psi for unmade-up connections with the V-038R thread form. Figure 5
shows the FEA modeling results, and the actual test data
obtained from Reference 8, at 0 MUT, and the maximum
allowable cyclic stress to obtain 10,000,000 cycles is about
15,000 psi.
An interesting outcome of the FEA modeling, and a topic
for additional research, was the affect of MUT on cyclic
stress-strain amplitude when MUT was low enough to keep
the root of the LET elastic. An increase in MUT (in the elastic
region) amplifies the cyclic stress-strain amplitude even when
the same bending moment (dog-leg) is applied to the
connection. The implication of this finding is that not only
does the mean stress increase with MUT (in the elastic
region), the cyclic stress-strain amplitude increases as well.

6

IADC/SPE 87190

It should be noted that the SLM and the actual testing
referenced in this paper does not take into consideration
dynamic loading such as downhole vibration and whirl. These
unknown variable forces make the calculation of an actual
endurance limit for RSC’s impractical, if not impossible.
However, what can be said is that modeling performed to date
suggests once enough MUT is applied to a RSC, additional
MUT has little effect on the fatigue life of the connection.
References 13 and 14 describe the concept of downhole
vibration, as well as the effects it has on the life of
downhole equipment.
To restate the factors that must be considered when
increasing MUT:

4 50 0 0
40000

Cyclic Stress Amplitude (psi)

3 50 0 0
30000
2 50 0 0
20000
150 0 0
10 0 0 0
50 0 0
0
1.E+0 3

1.E+0 4

1.E+0 5

1.E+0 6

1.E+0 7

1.E+0 8

1.E+0 9

1.E+10

C yc le s to F a ilure

0 ft-lb FEA-SLM

3.

Actual 0 M UT

Note: Actual data points obtained from Reference 8
Figure 5 - FEA Strain Life Modeling Results and Actual Fatigue
Test Results at 0 kft-lb MUT

At MUT near API specified (9 kft-lb), the SLM
predicted life vs. cyclic stress varied from the actual tests by a
wider margin than the 0 MUT case. Figure 6 shows the SLM
and actual fatigue tests with 9 kft-lb MUT on the NC-38
connection. The discrepancy is most likely due to differences
in the fatigue constants (material properties) used in the SLM
and the actual fatigue constants of the test specimens. See
Reference 12 for more detail on the fatigue constants that are
required for the SLM. Additional full scale testing should be
performed
to
correlate
the
SLM
with
actual
material specimens.

Cyclic Stress Amplitude (psi)

2 50 0 0

20000

150 0 0

10 0 0 0

50 0 0

0
1.E+0 3

1.E+0 4

1.E+0 5

1.E+0 6

1.E+0 7

1.E+0 8

1.E+0 9

1.E+10

C yc le s to F a ilure
9 kft -lb FEA-SLM

1.
2.

Actual 9 kft-lb M UT

Note: Actual data points obtained from Reference 8.
Figure 6. FEA Strain Life Modeling Results and Actual Fatigue
Test Results at 9 kft-lb MUT

4.
5.

Is the connection pin or box weak?
How does this additional tensile preload in the tool
joint effect the connection’s ability to carry tension?
What is the tool joint’s remaining tensile capacity at
the applied MUT?
What is the bearing stress at the sealing shoulder?
Will the fatigue life of the connection be reduced due
to increasing MUT?

The answer to question five, based on FEA modeling, is
that increased MUT will not significantly reduce the fatigue
life of the connection. However, extensive full-scale testing is
required to solidify this theory. The answers to the other four
questions above must be carefully evaluated and answered
before increasing MUT beyond the sound values stated in DS1TM, or API-RP7G.
Conclusions
1. The Strain Life Model suggests that the fatigue life of
a pin on a drill pipe tool joint RSC is not significantly
influenced by MUT (the mean stress), once enough
MUT is applied to plastically deform the root of
the LET.
2. The Modified Goodman Diagram is limited to the
elastic region and is not useful in predicting whether
or not a made-up tool joint, with plastic strain at the
root of the LET, will have infinite fatigue life or not.
3. Plastic deformation of the thread root occurs near and
above API MUT.
4. Cyclic stress amplitude, which occurs when rotating
pipe through a dog-leg, is the overriding factor when
attempting to predict fatigue life.
Nomenclature
T
= Torque (ft-lb)
A
= Cross sectional area (in2)
S
= Desired stress level from make-up
P
= Lead of threads (in)
Rt = Average mean radius of thread (in)
Rs = Mean shoulder radius (in)
f
= Thread compound coefficient of friction
Ф = ½ thread angle (Figure 9.2 or 9.3, API Spec. 7)
Tpr = Thread taper, on the diameter, (in/ft)
API = American Petroleum Institute
CLC = Combined Load Curve
COF = Coefficient of Friction

IADC/SPE 87190

FEA = Finite Element Analysis
IF = Internal Flush
KSI = Thousand Pounds Per Square Inch
LET = Last Engaged Thread
MD = Measured Depth
MUT = Make-up Torque
MYS = Minimum Yield Strength
NC = Numbered Connection
PSI = Pounds Per Square Inch
RSC = Rotary Shouldered Connection
SLM = Strain Life Model
TD = Total Depth
THHA = T H Hill Associates, Inc.
References
1. Chandler, R. B., Jellison, M.J., Payne, M.L., Shepard, J.S.:
“Performance Driven Drilling Tubular Technologies,” SPE
79872 (February 2003).
2. Jellison, M.J., Hassmann, S.P., Snapp, D.: “New
Developments in Drill Stem Rotary Shoulder Connections,”
SPE 62785 (September 2000).
3. Stuart, D., Hamer, C.D., Henderson, C., Gaynor, T., Chen,
D. C-K.: “New Drilling Technology Reduces Torque and
Drag by Drilling a Smooth and Straight Wellbore,” SPE
79919 (February 2003).
4. Baryshnikov, A., Ferrara, P., Schenato, A., Curioni, E.:
“Makeup Torque and Rotary Shouldered Connection
Reliability,” SPE 29352 (February 1995).
5. Winship, T.E., Vinson, B.: “An Investigation of Pressure
Capacity of Rotary Shouldered Connections,” SPE 39324
(March 1998).
6. Farr, A.: “Torque Requirements for Rotary Shouldered
Connections and Selection of Connections for Drill Collars”,
ASME Paper No. 57-Pet-19.
7. Dowling, Norman E., Mechanical Behavior of Materials:
Engineering Methods for Deformation, Fracture, and
Fatigue. Prentice Hall, 1993, ch. 9 & 14.
8. Bailey, E., Smith, J.: “The Goodman Diagram as an
Analytical Tool to Optimize Fatigue Life of Rotary
Shouldered Connections”, SPE/IADC 79874 (February
2003).
9. API RP7G, “Recommended Practice for Drill Stem Design
and Operating Limits”, sixteenth edition, API, Washington,
D.C. (1998).
10. T H Hill Associates, Inc.: Standard DS-1™ Drill String
Design and Inspection, second edition, (March 1998).
11. Lubinski, A.: “Maximum Permissible Dog-Legs in Rotary
Boreholes.” Journal of Petroleum Technology, (Feb. 1961)
p. 173.
12. Ellis, S., Reynolds, N., Lee, K.: “Use NC56 Connections on
8" Drill Collars and Cut 1" or 3/4" Pin Stress Relief Grooves
on Rotated BHA Connections NC38 and Larger,”
SPE/IADC 87191.
13. Shuttleworth, N., van Kerkoerle, E., Folmer, D., Foekema,
N.: “Revised Drilling Practices, VSS-MWD Tool
Successfully
Addresses
Bit/Drillstring
Vibrations”,
SPE/IADC 39314 (March 1998).
14. Ortiz, B., Casallas, C., Parra, H.: “Improved Bit Stability
Reduces Downhole Harmonics (Vibrations)”, IADC/SPE
36413 (September 1996).
15. Hill, T., Ellis, S., Lee, K., Reynolds, N.: “An Innovative
Approach To Reduce Drill String Fatigue”, SPE/IADC
87188

7