- Torque wrench calibration & training, friction variations, long learning curves, peak-jumping, ..., etc.
- Errors due to changes in lubrication or cross-threading.
The number one reason that bolted
flanges leak and bolts fail?
Pre-load tension errors. Why? Because a perfect torque wrench leaves
you with 20% to 30% tension
errors (greater errors on flange bolts). See
the graph below. How can you cure this? Answer: Use stress versus ultrasonic
strain, not stress versus torque.
Before we can talk about the solution, let us understand the source of
these tension errors.
when tightened to identical torque
values, will vary substantially in their actual tensions (typically plus or minus
25%, and not uncommonly plus or minus 50%
Friction Variations: Variation in tension happens because each bolt varies in the amount of
torsional work consumed by the
bolt's friction. Usually only about 10% to 15% of the torsional work
converts to tension in the bolt. Only 10% to 15%?
Where does the rest go?
And the friction varies so much, bolt
to bolt, that your final error is +-25% of your target tension and it is
not uncommon to have +-50% error (see
of errors), --- even with perfect torque control. (Friction variations cannot fool
This much error in pre-load
tension complicates the design of bolted joints, especially lightweight critical joints and flange
connections. Flange bolts vary even more and this variation is the primary reason some flanges
leak. See the graph below.
Calculations: If 85% to
90% of your torque is consumed by friction, then this only leaves 15% to
10% for actual bolt tension. 10% is 33% less than 15%. 15% - 10% = 5%,
but this 5%
difference in friction gives you a 33% variation in bolt tension.
And if you use a much better lubricant, then maybe
only 80% to 87% of the torque is consumed by friction. This leaves 20%
to 13% for actual bolt tension. 13% is 35% less than 20% (20-13=7.
Then 7/20 = 35%). So the 7% difference gives you a 35% variation in
tension. But, this reduction in torque - needed to get the same bolt
tension - leads to a 25% reduction in the required torque, but the variation in
friction still leaves you with a 35% variation in bolt tension. So the friction variation
is usually the same or worse, when you use better lubricants.
Email us the details of your bolted joint
application. So how accurately can ultrasonics be used? See Bickford
Using only a torque wrench, on flange bolts, can
large tension variations.
Notice that after 3 passes, bolt #3 had over 10 mils of tension, while
bolt #7 had less than 4 mils of tension. (10/4 = 250%
variation, or (10-4)/10 = 60%
variation. Either way the variation is
large. And bolt #15 is even worse.)
More passes did not help the
Torque pass, after pass, after pass. But look what happens --->>>>>
Notice the "typical" variation
pattern in bolt tensions on each of three passes with a
torque wrench. Their Y axis's "Stud Stretch
(mils)" shows the actual bolt tensions.
What if the extra tension (for example bolt #3) causes
your bolt to go into yield? Now you affect the bolt's fatigue life and
other safety factors.
X-axis = bolt position. 16
Y-axis = bolt stretch (tension).
Large tension variations are typical with
torque control. Notice the > 40% to 70% variations.
Notice John Bickford's statement: "The final sawtooth pattern is
of gasketed joints tightened under torque control."
If you use ultrasonics to do the final pass, the variations can be
greatly reduced. This is how many petrochemical companies fix their leaky
Give the credit to
John Bickford for this graph.
Mr. Bickford is probably the world's expert on bolted joints.
Read below to see how to measure bolt tension using
said: " Using a digital signal processing (DSP) technique
(patented), the gage (SureBolt) proved more reliable on every test and
bolt than any other available gage."Download
the NASA story (pdf file, 264k).
Update says: " SureBolt outperformed the one-point bolt gauges
on every bolt and every test in reliability and accuracy."
Thread lubrication allows more of
the torque to be applied to the bolt's tension, but the bolt to bolt tension
is still the same, and sometimes worse.
SureBoltTM does not get fooled by friction.
There are many techniques for reducing
these frictional errors but the use of an ultrasonic bolt gage allows the actual
bolt tension to be measured more directly --- in particular, the ultrasonic pulse is not affected by the bolt's friction losses.
if you think calculating the correct torque is easy, just see either of these
two web pages (no affiliation with FAST-DAQ), nut
factors or torque is
With SureBoltTM, instead of calculating the correct torque, you just measure the tension
You can even use
SureBoltTM to determine the correct torque and its variation. How does this work?
said: "We have finalized our investigation of the UT (ultrasonic
currently available on the market and we would like to purchase the
Surebolt Gage." This is
why we give you the links to our competition. Our competition is our
Tension, not torque. Accuracy and reliability.
SureBoltTM can be
thought of as a super accurate ruler (micrometer) that measures how much a bolt
stretches under load (sometimes within
+/- 0.00001 inch or better). How?
SureBoltTM sends a sound pulse (ultrasonic) into one
end of a bolt, and then measures the time for the echo to return (like sonar). When
you tension the bolt, the echo takes longer to return! See
correlation bolt gauge was first developed for NASA's Space Station. SureBoltTM
takes most of the guesswork and frustration out of ultrasonic bolt gage use. See
why NASA switched to SureBolt.
Competing ultrasonic bolt gages track only one point on the echo. The
echo often (10% to 30% of the time) distorts when the bolt is tensioned. This distortion of the echo
- which can destroy the performance reliability of competing gages - has little affect
on SureBoltTM. See
This is because SureBoltTM
uses all the points of the waveform, not just one zero crossing
point, and is therefore not as prone to the infamous problem of "peak
jumping" (see graphics).
Read the "In Conclusion" section below and all the problems that are
"GONE", when you use
SureBoltTM. Some case
histories follow the conclusion section.
Tell us about your bolted joint application.Email us, write
us, or phone us. We may already have an answer. Or maybe between both of us, we
can come up with another solution. Purchase
SureBolt or lease with option to buy.
Use of a one-point ultrasonic bolt gage will
give you far more accuracy than a torque wrench, if you
somehow know when you can rely on its reading.
A new technology developed for NASA's Space
Station is now available. NASA needed accurate tension data on very
critical Space Station bolts (see NASA story).
This patented technology
uses the entire echo instead of just one point (see
Compared to competing bolt gages, the calibration process is tremendously simplified and the user
interface is the familiar Windows 2000/XP (buttons, graphs, sliders, pictures...).
Click here for a simple explanation of how SureBoltTM
works and see a simulation of the difference between SureBoltTM
and all other bolt-gages.
In addition, SureBolt gives you a fully loaded,
Panasonic Toughbook laptop computer (13.3" TFT color
uses Windows 2000/XP to make your learning experience easy. You will not have to
suffer with short file names, buttons with multiple uses, small black and white screens, ... and
SureBolt uses industry standard hardware/software and operating
has a longer useful upgradeable life.
GONE - 25%
to 50% tension errors.
GONE - Torque wrench training,
procedure errors, special washers, friction variations, lubrication
errors, ..., etc.
GONE - Large
variations in tension on typical flange bolts. Even with many passes, and
tensioning the bolts in a special order, flange bolts have large tension
variations if torque or turn of the nut methods are used. Multiple passes
do not help.
GONE - Many extra torque passes, star patterns, and
many procedures for flange bolts.
GONE - Long learning curves,
cryptic bolt names, small B/W screens, limited keyboards, peak-jumping,
GONE - ASTM
procedure that says you must calibrate your torque wrench daily.
GONE - Defining
"snug-tight" condition for the "turn-of-the-nut"
GONE - Bolts with less than 1/2
turn to fully tightened have large errors when using the turn-of-the-nut
GONE - Torque wrench
regulation that says
"Installation procedures shall be calibrated at least once per day,
on 3 or more bolts, to determine the actual bolt tension.
Proprietary operating system. Totally proprietary hardware. Questionable
life expectancy due to total dependence upon the manufacturer's limited
Requirement for hardened washer under the turning element (nut or head).
Case Histories from "An
Introduction To The Design And Behavior Of Bolted Joints (3rd ed. by
John Bickford) Probably the best bolted joint resource available. Again,
no affiliation with SureBolt or FAST-DAQ.
(pg 410) The additional examples below will help
illustrate the fact that ultrasonics allows you to see things you cannot
see as conveniently by any other means. In fact, I think it gives the
bolt mechanic or engineer the same sort of insight the electrician gets
from an oscilloscope. The mechanic can make a change in lubricant,
tools, fastener material, etc. , and, thanks to the bolt gage, see
immediately the results which that change has produced -- instant
feedback in either a static or a dynamic situation. For the first time
we can go into the field and determine actual results on actual bolted
joints -- not just on a few laboratory or field samples specially
prepared with strain gages or the like.
The things you will see will not always be pleasant
or understandable. Bolted joints are a very complex subject, and their
behavior is not well known. They are not very stable, either-- they flex
and shift and relax under environmental changes and/or variations in
external load--as well as just "all by themselves." When you
start to use ultrasonics, therefore, you should be prepared for a
number of unpleasant surprises. ... Laplace who said, "Getting an
education is like climbing a flagpole; the higher you go the more
horizon you can see." Using ultrasonics on bolted joints gives you
such an education. Here's a typical example.
A. Connecting=Rod Bearing Caps on a Large Diesel
The fasteners were 1 3/4 - 10 x 17, 4140
studs. The problem: Fatigue failure of one stud had created a chain of
events which lead to the failure of the entire engine. Ultrasonics was
used to answer some of the questions raised by the analyst.
First question: What would failure of one stud do
to the load distribution in the other seven studs holding down the
To answer this question, eight new studs were used
to fasten a bearing cap onto an actual engine. Ultrasonics was used to
control the elongation in each. One stud--that corresponding to the one
which had initiated the original failure--was now loosened completely.
The residual preload in the remaining studs was measured, with the
results shown in Fig 11.13.
Second question: How much residual preload is there
in a typical bearing cap stud 1.5 years after initial tightening? The
engine has been running most of this time.
This question was answered by measuring the loss
in length(bold not in original)
in 16 studs (two caps in another engine) when they were removed for
routine maintenance purposes. ... These results were somewhat surprising
In any event, use of ultrasonics gave the analyst
some joint behavior and joint condition information that would have been
difficult or impossible to obtain by any other means.
Other valuable information.
See how NASA used SureBolt's Whole Echo patented
technology to greatly increase reliability, accuracy, and ease of use.
other bolt gages use just "One Point" (one zero crossing), that leaves
you susceptible to peak jumping (30% or more error). SureBolt is a more
reliable bolt gage due to this new patented technology. See
product comparison table.
Bolt gages have been around for over 20 years. Their "Peak
Jumping" problems and long learning curves, have kept them from being
When your bolted joint fails or flange leaks, then
you find out your one-point bolt gage may have jumped peaks. And you still
cannot be sure, because you only recorded "one-point", not the whole
With a lot of training and laboratory testing, NASA has used bolt gages for
years. NASA physically takes a Polaroid picture of the echo before and after
Then they use their trained human experience to determine if the
echo has distorted "too-much". Even after all this, they have to throw
away approximately one third of the ET (External Tank umbilical bolts)
measurements because they are not reliable.
And if you think calculating the correct torque is easy, just see either of
these two web pages (no affiliation with FAST-DAQ or any bolt gage manufacturer), nut
factors or torque is
With SureBolt, instead of calculating the torque to get you the
correct tension, you just measure the tension more directly. SureBolt is
like a strain gage, only SureBolt measures the strain in time (echo return
You can even use
SureBolt to determine the correct torque, and its tension variations (friction
variations, different procedures, different lubricants, ... etc.).
SureBolt gives you a 13.3" TFT touchscreen, portable PC
(Panasonic Toughbook Model 72). The bolt gage hardware is built into the CD ROM and PCMCIA
slots. Click here to see the actual SureBolt
See the animated DEMO page to see how
an ultrasonic bolt gage can measure tension. SureBolt has been used on a wide
variation of bolts from special
high strength titanium bolts to standard "off the shelf" bolts. SureBolt handles more bolt types than any other bolt gage.
Quotes from "An Introduction To
The Design And Behavior Of Bolted Joints (3rd ed. by John Bickford) Probably
the best bolted joint resource available. No
affiliation with SureBolt or FAST-DAQ.
(pg 373) Measuring or controlling bolt stress or
strain ultrasonically is not uncommon at the present time. The equipment
and techniques are relatively new, ... ultrasonic equipment is
used primarily in applications involving relatively few bolts in
critically important joints, or is used for quality control audits, or
to set up and calibrate other types of assembly equipment (torque,
torque-turn, etc.), or to conduct laboratory or field experiments.
Currently available ultrasonic equipment is ideal
for the QC (Quality Control) and laboratory roles because it can monitor
bolt tension not only as it is being developed in a bolt during
assembly, but at anytime afterward. It can measure residual preloads
after embedment or elastic interactions, it can observe changes in bolt
tension as external loads are applied, it can monitor the effects of
temperature change, etc. It's a low-cost alternative to strain-gaged
bolts in all sorts of studies -- "low cost" because such bolts
can cost several hundred dollars each.
Currently available instruments can be used on
virtually any kind of bolt material, including steel, aluminum,
titanium, Inconel, MP35N, and other exotic materials. There are however,
some practical limitations ...
(pg 403) ... As a result, we feel that the
equipment is capable of residual preload measurements with an accuracy
approaching +-5% in most field situations, and can approach +-2% in
laboratory work. There are even well-documented examples where +-1% was
achieved consistently in aerospace applications; but ...
See our technical information page for Barnes
& Nobles links to where you can buy Mr. Bickford's books. Again, Mr.
Bickford has no affiliation with SureBolt.
Side Note: SureBoltTM is
built into a Panasonic ToughBook (pdf flyer 817k). SureBoltTM
uses industry standard hardware & software for a long support life, and
allows you to have the familiar PC user interface.
Your computer becomes the instrument.
the CD bay and PCMCIA slots.
Tell us about your application.
Send your e-mail to with
questions or comments about how we can help. If you want to contact us,
please use this contact page, or type in the email address shown above (due to
email spam, we have replaced our email address text with the image shown above).
Note: We respect your privacy. We will NOT give your address to anyone, for any
Number of visitors
to this page, since 1-18-2002
FAST-DAQ Let us know how we can help you.
1022 Antelope Trail, Winter Springs, FL 32708, USA
Send your email to
comments about how we can help. Contact us here,
or type in the email address shown above (Due to spam, we have replaced the text
email address, with the image email address shown above).
Note: Privacy. We will NOT give your contact information to
No more guessing as to
which "one point" is the right point.
The FIRST whole
echo method (patented DSP Technique).
Built into a
Visitors Since 1-22-02
September 27, 2011