Strain gauge measurements of friction on radial dam gate bearings

Henning Føsker, Halvard Bjørndal & Terje Ellefsrød

Norconsult AS, Sandvika, Norway
Kjell Knutsen
Sira-Kvina kraftselskap, Tonstad, Norway

ABSTRACT: At least five known radial gate failures are caused by trunnion bearing seizure. The radial gate
arms are normally designed to withstand bending moments from nominal friction on the bearings. Experience
has shown that lack of lubrication and years of deterioration leads to increased friction and even seizure of the
bearings. The bending moments produced by bearing friction imposed on the gate arms are beyond the mo-
ment capacity, resulting in collapse. Norconsult has developed a method for measuring trunnion bearing fric-
tion while the gates are in service. By means of strain gauges attached to the gate arms, the strain caused by
the bearing friction is measured during gate movement. The signals from the strain gauges and a gate position
transmitter are logged digitally, allowing direct presentation of test results. The method gives objective and
precise verification of measuring results, and the risk of subjective and wrong assessments is minimized.

1 BACKGROUND an evenly distributed load, and the grease escapes

through the low-pressure side of the radial bearing.
1.1 Introduction
The bearing material and opposing surface often
Increased friction and seizure of bearings on dam represent an unfavorable material combination be-
gates is a growing concern. At least five radial gate coming corroded and contaminated. This is probably
breakages in Norway and abroad are caused by bear- the most important factor in deterioration of bear-
ing failure. Consequently, there is uncertainty ings. When water is present and an electrolyte can
among dam owners regarding the state of the bear- occur with the carbon steel shaft as the anode. This
ings on radial gates after several years of service un- leads to corrosion on the bearing surface of the shaft
der severe conditions and sparse maintenance. and increased roughness from pitting and corrosion
products. Tight bearing radial clearance worsens this
effect.
1.2 Deterioration of bearings
Early dam gates (before 1980) are most commonly
1.3 Damage mechanisms and consequences
designed with trunnion bearings using a carbon steel
shaft and bronze radial and axial bearings. The Radial gate arms are often slender steel structure
shafts are often chromium plated, and the bearings dimensioned to withstand bending and buckling
are usually lubricated by a manual grease lubrication from the water pressure. Friction forces are often not
system. taken into consideration, or nominal values of fric-
Dam gate bearings are static bearings with small tion coefficient are used. Normally a lubricated
rotation and relatively few movements during opera- bronze bearing and carbon steel shaft has a nominal
tion. This operation mode is not suitable for a bear- friction coefficient of µ = 0.2 – 0.3
ing requiring external lubrication as the lubricant is Deterioration of the bearings leads to increased
not distributed on the high pressure sector of the friction, for which the arms are not dimensioned.
bearing. When the increased friction causes breakage, this is
When injected to the radial bearing, the lubricant due to both fatigue and instantaneous breakage. In
often finds its way to the low-pressure side, leaving particular, cylindrical bearings with a manual lubri-
the high-pressure load transformation sector unlubri- cation system combined with an automatic or re-
cated. The axial bearings are often connected to the motely operated gate are a vulnerable configuration.
same lubricant pipe as the radial bearing. This gives All radial gate collapses lead to economic loss
due to loss of power production. Excessive dis-
charge results in downstream flooding and lowering 2 EVALUATION OF METHODS
of the reservoir, affecting boat traffic and recrea-
tional use of the water way. 2.1 Inspection and maintenance
Inspection and maintenance of gate trunnion bear-
ings often implies dewatering of the dam gate and
1.4 Arm design
dismantling of the bearing. Dewatering of the gate
Exposure to failure due to increased bearing friction can be accomplished by lowering the reservoir. This
depends on the design of the arms. Areas of concern is, however, both time-consuming and costly if pro-
are the transition zones between the trussed beams duction is lost. More common is dewatering by in-
constituting the arms and the hub accommodating stallation of stop logs in front of the gate. This is less
the bearings. As an example we have analyzed the expensive than lowering the reservoir, but still more
effect of a friction coefficient of µ = 0.5 on the Lun- expensive than other methods. Another disadvantage
devann dam radial gate arm. At full water head it is that during the de-commissioning period the dis-
imposes a bending moment resulting in a bending charge capacity of the gate is not available. Disman-
stress of 7MPa in the large cross-section hub. The tling of radial gate trunnion bearings is subsequently
bending moment is transferred to the more slender not routinely carried out to determine the condition
trussed beams, constituting the gate arms. In the of the bearing. Dewatering and dismantling of the
transition zone between hub and arms, the stress bearing is carried out when and if the symptoms of
concentration is especially high in this example. A bearing seizure are evident. On some radial gates to-
friction coefficient of µ = 0.5 results in a bending tal collapse has occurred without prior symptoms.
stress in the transition zone of 100MPa as shown in
Figure 1. This stress change goes from plus to minus
2.2 Detection of change in maneuvering forces
when the gate movement changes direction, expos-
ing the transition zone to fatigue. In theory an increase in lifting force could be de-
tected when bearing friction increases. However, the
bearing friction constitutes only a small fraction of
+100Mpa the total lifting force. On a typical radial gate with
upstream lifting chains, the total lifting force of
156kN is dominated by the gates own weight of
0Mpa 126kN (81%). The remaining 30kN (19%) is friction
in rubber seals and trunnion bearings. To separate
-100Mpa the two friction forces, the friction on the seals must
be based on an assumption regarding the friction co-
efficient between rubber seals and the embedded
stainless steel frame. Assuming µRUBBER-STEEL = 0.9,
the friction on the seals constitutes 20kN (13%) of
the lifting force. The remaining, 10kN (6%) is the
Figure 1. Lundevann dam radial gate arm, hub section. Stress
nominal bearing friction based on µBRONZE-STEEL =
resulting from friction coefficient, µBRONZE-STEEL = 0.5 0.2. A variance of this relatively small force, com-
pared to the total lifting force, is complicated to de-
tect and is unreliable since the method is based on
1.5 Trunnion design assumptions.
Trunnion designs vary with regards to vulnerability
for bearing seizure. For the common cantilevered 2.3 Detecting bending moments in radial gate arms
trunnion shaft anchored to the abutment, symptoms
of bearing seizure are difficult or impossible to de- As a part of a safety revaluation program for dams, a
tect by visual inspection. High friction moments can diagnostic technique has been developed for radial
be transferred through the shaft without visible indi- gate bearings without dewatering the gate or disman-
cations of increased bearing friction before the col- tling the bearings. The method detects friction on the
lapse occurs. bearings during operation of the gate.
The trunnions may include a shaft lock allowing
rotation of the shaft in the console, causing the shaft 2.4 Measurement technique
lock to be deformed after a bearing seizure.
The forces acting on the trunnion bearing can be di-
vided into two classes:
 Perpendicular forces on the bearing surfaces is
the result of water pressure, gate weight, operat-
ing forces and friction forces from the rubber
 seals. The resulting force from the water is the strain caused by thermal changes is compensated to
dominating the force on the trunnion. a high degree.
 Shear forces parallel to the bearing surface due to
friction. Without friction (µ = 0), no shear forces
will appear. These forces will give bending mo- 12 Measured bending stress 1.5
ments in the gate arms.
Gate opening

Bending stress [MPa]

Gate opening [m]

8
1.0

0.5
0

-4 0.0
0 200 400 600 800
Time [s]
Strain gauges
Figure 4. Lundevann dam, measured stress in right gate arm.

Figure 4 shows the measured stress in the right gate

Figure 2. Strain gauges attached to the radial gate arms, close arm at the Lundevann dam with the original bearing
to the bearing. thoroughly lubricated. The gate was opened, almost
closed, opened and finally closed. The vertical gate
The shear forces on the bearings cannot be measured opening was measured simultaneously as shown in
directly without modifying the gate structure. In- the figure.
stead the mechanical stress variation is measured in
the gate arm, near the trunnion bearing, using strain
12
gauges. One strain gauge is attached at the upper
Measured bending stress [MPa]

side of the gate arm, and one is attached at the lower Stress variation due
side of the gate arm as indicated in Figure 2. Each 8
to bearing friction
strain gauge measures the surface mechanical stress
in parallel to the main stress direction in the gate
arm. 4

0
M easured strain by Strain d u e Strain d u e to
to ax ial lo ad b end in g load Mean stress
strain gau ge #1
-4
0.0 0.5 1.0 1.5
= + Figure 5. Mean stress variation.
Gate opening [m]

M easured strain by Strain g aug e # 1

strain gau ge #2 - Strain g au ge #2 The measured stress caused by bending is shown at
the Y-axis and the vertical gate opening at the X-
axis, see Figure 5. The figure shows that the meas-
2 * Strain d ue to
bend ing lo ad ured stress creates a hysteresis curve superimposed
on a mean stress variation. The variation in mean
Figure 3. Stress measured by strain gauges. bending stress is mainly caused by variation in how
the flow forces act on the open gate. These forces
Superimposed axial and bending forces cause the are dependent only on the gate opening, not the di-
measured stress indicated in Figure 3. By connecting rection of the movement. The variation of these
the strain gauges in a Wheatstone half bridge circuit forces is not used in the analysis and may be ig-
the two measured stresses are subtracted and the nored. The friction forces in the bearing always act
output only indicates active bending moment in the against the direction of the gate motion. When the
gate arm (ref /1/). By using this circuit, superim- motion changes direction the measured bending
posed axial (normal) strain is compensated, and the stress, due to friction forces in the bearing, also
changes direction. In addition the friction force var-
ies with the gate opening, as the water load on the 3 MEASUREMENTS ACCOMPLISHED
gate alters.
Dependent on the hoist design, the friction forces 3.1 General
from the gate rubber seal have some influence on the Since the method was developed in 1998, a total of
total measured friction force. Upstream chain hoist 30 bearings on 15 radial gates in six different dams
has no influence. In this actual design the rubber seal have been measured. On one of the dams where in-
friction gives less than 5% effect on the measured creased bearing friction was detected, Lundevann
bending stress when we assume µRUBBER-STEEL = 0.9 Dam (owner: Sira-Kvina Kraftselskap), repeated
(worst case). This estimation is compensated in the measurements were carried out .
calculations.
The friction torque at the bearing trunnion is cal-
3.2 Presentation of results
culated based on the measured bending stress due to
friction. This calculation is done as a function of Table 1. Key data of gates measured and results
moment of inertia in the gate arm at strain gauge lo- WxH Bearing Lubr. Friction coefficient Prior
cation and radial distance from the strain gauges to m Shaft system Nomi Measured lubri-
the bearing centre line. nal Left / Right cation
The water load on the gate is calculated as load 15 x 5.2 B/S Grease 0.2 0.08 / 0.09 Yes
from the static water pressure at the wet surface of 8.0 x 2.8 B/SS Selflub. 0.1 0.35./.0.25 N/A
the gate. The accuracy of this calculated load is high 20 x 8 B/PS Selflub. 0.1 0.15 / 0.10 N/A
when the gate is closed. At increased gate opening, 13 x 4 B/S Grease 0.2 0.75 / 0.50 No
larger areas are exposed to flow velocity, which de- 13 x 4 B/S Grease 0.2 0.45 / 0.45 No
crease the static pressure. Our method will over- 11.5 x 5 B/S Grease 0.2 0.15 / 0.14 Yes
estimate the load when the gate opening increase, 11.5 x 5 B/S Grease 0.2 0.14 / 0.13 Yes
but according to pressure distribution calculation on 11.5 x 5 B/S Grease 0.2 0.12 / 0.09 Yes
a gate (ref /2/) this error will be limited to approxi- 13 x 4 B/S Grease 0.2 0.55 / 0.40 Yes
mately 15 % at 25% gate opening. This calculation 13 x 4 B/S Grease 0.2 0.22 / 0.35 Yes
gives the main input to the radial load at the bearings 12 x 6.3 DU/PS Selflub. 0.15 0.18 / 0.20 N/A
transmitted through the gate arms. 12 x 6.3 DU/PS Selflub. 0.15 0.17 / 0.19 N/A
Dependent on the design, the gate arm axial force 17 x 5 B*/PS Selflub. 0.1 0.71 / 0.58 N/A
is split into two components, one acting on the axial 17 x 5 B/PS Selflub. 0.1 0.43 / 0.42 N/A
bearings and the second on the radial bearings. This 13 x 4 C/SS Selflub. 0.15 0.09 / 0.13 N/A
split is dependent on the gate design. When both ra- B = Bronze, S = Carbon steel, PS = Chrome plated carbon
dial load and friction torque on the trunnion bearings steel, SS = Stainless steel, DU = Glacier DU bushing, C =
composite. * Oiles 500 w/SL4 lubricant plugs.
are known, we can calculate the average friction co-
efficient on the bearings. Figure 6 shows the friction
coefficient dependent on the gate opening and direc- 3.3 Lundevann Dam
tion of motion.
The Lundevann Dam is the reservoir for the Åna-
Sira hydropower plant. In the outlet of Lundevann a
concrete dam was constructed with two spillways
0.5 Friction coefficient 2.5 and surface radial gates.
Gate opening
Friction coefficient [ - ]

0.4 2.0 Table 2. Key data for the radial gates on the Lundevann Dam
Gate opening [m]

BxH : 13.0m x 4.2m

0.3 1.5 Reservoir level : El. 48.5
Bottom sill : El. 44.5
Gate radius : 7000mm
0.2 1.0 Old radial bearing diameter : 250mm
Old axial bearing diameter : 370mm
0.1 0.5 Old bearing material: : Cast steel shaft/bronze bush
New radial bearing diameter : 216mm
0 0.0 New axial bearing diameter : 350mm
New bearing material : Stainless steel/Orkot**
0 200 400 600 800
** Trade Mark from Busak + Shamban, non-metallic self-
Time [s] lubricating bearing material
Figure 6. Calculated friction coefficient

The measurements were carried out on the bearings

under three different conditions:
1 Original bearing without prior lubrication 3.3.4 Comparison of the measurements
2 Original bearing thoroughly lubricated Figure 7 shows the results from three of the meas-
3 New bearing with self-lubricating bushing urements of the right bearing on gate No. 1 on Lun-
devann dam.

3.3.1 Measuring the original bearings

The first measurements on the four bearings, two on
0.8 Original bearing
each gate, were carried out in 1998. Prior to the
0.7 Original bearing, lubricated
measurements, the gates had been stationary for one New bearing

Friction coefficient [ - ]
year without maintenance or lubrication. The meas- 0.6
urements revealed increased friction values in all 0.5
bearings. The friction on the bearings of the most
0.4
frequently used gate No. 1 was especially high. On
the first opening movement, the friction coefficient 0.3
was 0.75, dropping to left 0.60/ right 0.50 in the sec- 0.2
ond opening. The coefficient of friction remained 0.1
considerably higher than expected for the material
properties of the bearings. 0
In 1999, we repeated the measurements after the 0.0 0.1 0.2 0.3 0.4
bearings had been thoroughly lubricated and moved Gate opening [m]
repeatedly to distribute the grease. The friction coef- Figure 7. Comparison of measurements
ficient then fell by 25% to 0.45 and 0.35 on the same
bearings. Our conclusion was that the bearings had The initial (static) and dynamic friction of the origi-
suffered permanent damage and we recommended nal unlubricated bearings is almost constant. When
dewatering and replacement. lubricated the original bearing has almost the same
static friction as unlubricated, while the dynamic
3.3.2 Inspection and replacement of the bearings friction drops by 25%.
In 2000, the gates were dewatered and the bearings The uneven shape of the graph for the original
were dismantled. The inspection revealed traces of bearing indicates increased roughness of the bearing
seizure on the bronze surfaces and seizure and cor- surfaces, while the even shape of the graph for the
rosion on the surfaces of the cast carbon steel shaft. new bearing indicates smooth bearing surfaces.
It was also evident that the manual lubrication sys-
tem had not functioned as intended. Partly due to
clogged grease channels and partly due to the fact 4 CONCLUSION
that the grease escapes from the bearing on the low
pressure side. The method of measuring bearing friction by means
The bearings were completely replaced. New of strain gauges provides dam owners with a better
shafts were manufactured in stainless steel, SIS diagnostic technique. Experience has shown that the
2387. For radial and axial bearing, a non-metallic method will detect bearing failure at an early stage
material was chosen. Orkot® is a brand name from before the friction moment exceeds the gate arms
Busak+Shamban. Orkot® TLM Marine grades are bending moment capacity. The method is character-
non-asbestos composite materials incorporating ized by high reliability and accuracy. The influence
woven fabric reinforcement and solid lubricants from dynamic conditions and the friction on the
within a thermosetting resin matrix. The manufac- seals is insignificant. Measurements are carried out
turer gives a friction coefficient between 0.10 and without dewatering or decommissioning the gates.
0.15 against stainless steel. The method gives objective and precise verification
of measuring results, minimizing the possibility of
3.3.3 Measuring the replaced bearings subjective and wrong assessments.
After assembly and commissioning of the new bear-
ings, the friction on bearings on gate no. 1 was
measured. The result shows that the friction coeffi- REFERENCES
cients are within the values given by the manufac-
turer. The measured friction coefficient is 0.09 and Ref /1/ Hoffmann, Karl. An Introduction to Measurements us-
0.13 for the left and right bearings respectively. The ing Strain Gauges. 1989, Darmstadt: Hottinger Baldwin
results from the measurements of the new bearings Messtechnik GmBh.
Ref /2/ Wickert, G and Scmausser, G. 1971. Stahlwasserbau.
with the known properties and characteristics of the Berlin: Springer Verlag.
bearing materials is also a verification of the diag-
nostic method.