Professional Documents
Culture Documents
BI-DIRECTIONAL SAFETY
GEARS FOR CABIN
by Bertoni Stefano, Technical Director, Montanari Giulio & Co.
The recent introduction in Europe of the directive 95/16/EC indicate the sum total of that mass but rather to list in
and of the corresponding harmonization standard EN 81, detail the mass of the capacity (Q kg), the mass of the
have brought with them numerous changes in the elevator cabin plus the frame complete with accessories (P kg)
field thereby revolutionizing a way of thinking entrenched and the mass of the counterweights (Mcwt kg). These
after decades of stagnation. Among the numerous variations figures are the minimum required to verify the suitability
introduced in the area of safety, one of the more impor- of the desired equipment, whether dealing with a safety
tant ones is, without doubt, the obligation to furnish the gear or another device.
elevators with gear designed to prevent uncontrolled move- In fact, it’s necessary to be able to calculate the force
ments of the cabin when traveling upwards. necessary to brake the empty cabin traveling upward in
The EN 81 standard sanctions the minimum safety criteria order to then compare it with the characteristics normally
Elevator Safeties and Governors
ELEVATOR SAFETY
AND GOVERNORS
by K. Subramaniam, Johnson Lifts Private Ltd., Chennai, India
between guide and safety gear roller or cams are to be In older lifts, the gover-
checked and gaps must be uniform on either side and on nor wheel will have gun-
both ends. By adjusting the guide shoes and the safety, metal or bronze bushing
the required clearances can be obtained. In the older cam and similar bushings will
design, the rope to the master cam should neither be too also be there in the spring
tight nor too slack. For such cam designs, normally the rope guides, weight, pivots, etc.
is from a spring-loaded arrangement known as “torpedo Over a period of usage,
rope release.” This torpedo rope release will avoid actuation these bushings are likely to
of cams due to any jerk in the car. The spring in the torpedo wear out, creating noise
release must be to the correct pressure. The same procedure during lift running and if
is adopted for the gradual safety gears as well. left unattended will start
The rope from the governor to the car anchorage must getting actuated even at
be in plumb from where it is connected to the safety gear. normal speed. This noise
will be one of the prompting
In older designs, the return rope from the tension weight
factors to replace the bush-
is anchored separately at the bottom of the car. The rope
ing. Being calibrated equip-
from the governor to the tension weight must also be in
ment, much care is to be
plumb and adjustments will normally be available in the Torpedo rope releaser taken to recalibrate the
tension weight brackets to make good any error in the
governor after the necessary replacement and repairs of
plumb line. Care must be taken to ensure that governor the bushings. Preferably, the recalibration should be done
Elevator Safeties and Governors
ropes are free from any obstruction along the travel of the on a regular test bench and not on the installation site. In
lift. There are occasions where locating the governor and governors with weights, there are chances that oil with
tension weight becomes a difficult task, especially in cramped dust can clog the bushings preventing free movement
lift shafts. The governor rope holes in machine room floor of the weights. This particular aspect is to be checked
slab to be made good without any undulations and projec- periodically, preferably at every routine service call,
tions, preferably to be finished by inserting a two-inch especially in installations with dusty atmospheres and
PVC pipe to the full thickness of the floor slab. very cold places.
Any mistake in safety gear location with respect to the The same could happen even to the safety gear assembly.
guides could lead to severe damages to the guides. When- Over a period of time, governor rope tends to stretch and
ever safety gear is actuated if gripping is not uniform, the elongate, especially in very tall buildings. With this stretch
impact will tend to create a twisting moment that can result of ropes, the tension weight goes down. Once it reaches its
in the bending of the guides. limits, the rope becomes slack, and chances are it can get
entangled causing severe damages.
Nowadays, many governors are made with a test groove
An electrical safety switch is provided for such elongation
of smaller diameter. Transferring the rope to the test groove
of the ropes, but many times, you will find it bypassed.
will create overspeed for the governor with lift running at
This is done at the first instance as a temporary measure
normal speed and the safety gear can be tested. It is recom-
and seldom gets rectified on subsequent visits. Mainte-
mended having the safety gear test done with a rated load
nance personnel should check the switch and the status of
in the car. the tension weight and necessary corrective action be
Guide Lines on Maintenance and Repairs planned and completed at the earliest. In certain buildings,
Unlike the other components in an elevator, the safety rodents like rats getting in between rope and wheel can
gear and the governor are passive components and do also lead to such damage as the rope getting entangled.
not perform in the normal working of the lift. Even in the
governor, except the wheel rotating with the movement
of the car, other elements are passive. Normally, less
attention is given to the safety gear and governor by the
maintenance personnel and is mostly taken for granted.
Though much attention need not be given during the
routine maintenance visits, nevertheless it cannot be
totally neglected or ignored. It is advisable to check the
safety gear linkages at least once in four months and look
for problems in pivots and joints and also to check the
system for free movement. Typical safety gear link mechanism
This method has a number of advantages. The most ation due to gravity, 9.81mps2.
important of which is that it prevents any damage to the 3. The force of the safety gear: This applies in the upward di-
lift system components caused by the large forces generated rection, as it is a frictional force resisting the movement of the
during a full-load application of the safety gear. car in the down direction. It will be denoted by FSG in Newtons.
4. The upward force in the ropes acting on the car: Assuming
This article describes a number of practical tests carried
that the ropes are still in tension during the application of the
out to validate the method for heavy duty lifts. The relevant
safety gear, then they would apply an upward force. This force
formulae are derived and analyzed. The conditions under
is the resultant of two forces: The weight of the counterweight
which the method applies are also discussed.
(denoted as FC/W), and the traction force from the sheave re-
Introduction
sulting from the inertia of the drive (denoted as FT).
One of the most onerous requirements of the LG1 is the
The resultant force on the car would be: Resultant
carrying out of a full-load rated speed test on the safety
Force = FSG -FP/L-FC + FC/W - FT Continued S
gear. There is always concern about the damage that
could take place to lifts from such a test. The so-called
TÜV method uses a software tool that allows prediction Inertia of
rotating masses
of the results of the fully loaded test based on measurements FT
during a no-load test. LG1 in fact does allow alteration to
some tests, provided a risk assessment is carried out:
“In order to determine which examinations and tests should
be carried out and with what frequency, the installation
in question should be the subject of a risk assessment FC/W -FT
FC/W
that includes consideration of design, condition, usage of the
lift, relevant component manufacturer’s recommendations.
The results of the risk assessment may necessitate variations Car
The derivation of the general case equation is contained 3. The third assumption, which the method makes when
in Appendix B. The derivation of the two special case testing higher speed lifts, is that the safety gear force (fric-
equations is contained in Appendix C. These equations tional force) is independent of speed. This is not completely
are necessary in order to understand the derivation of the true [5]. Although friction is broadly constant, there is
so-called TÜV method, introduced in the next section, nevertheless some dependence on speed.
and which is pivotal to the argument of this method. 4. The fourth point is not so much an assumption as much as
The TÜV Method Masses and Its Application it is a simplification for measurement purposes. The method
In 1990, the TÜV in Germany developed a system called when reading off the deceleration from recorded graph, tries to
the ADIASYSTEM [1,2,3 & 4], which is basically a computer- “fit” a best “straight” line to the deceleration curve.
aided system for lift testing. Although the core of the system The last point becomes clearer when looking at a sample
is the measurement of the deceleration caused by safety gear curve taken from TÜV literature [2]. The curve in Figure 2
application, it also has some other features. These include, shows logged data of the acceleration in the vertical direc-
among others, measurement of traction and door pressure.
tion, using a data logger. The acceleration is integrated to
The method attempts to do away with the use of loads
produce a speed curve. The operator then has to move two
during the safety gear test. It carries out a measurement
vertical cursors on the computer screen, to identify the
during no-load application, and infers what the value at
start and the end of the safety gear operation (denoted as
free fall full-load would be. The method was formally approved
the upper and lower limits respectively in the figure). The
by the German Lift Committee in 1992, and is now used
cursors are placed so that they fit the best line of decelera-
in about six European countries [4] and has been trialed in
tion. Once that has happened, the software automatically
Canada [7]. Up to June 1999, 650,000 lifts had been tested
gives us the important equation Equation 6 derived in The important word to note here is free fall. A full-load
Appendix C): test at rated speed with the suspension intact does not nec-
aNL – g x Rm essarily provide the worst case scenario. So there is an
aFFFL = ––––––––––––– (6) extra benefit and further assurance using the TÜV method.
1 + Rm Site Test Results
mP/L In this section, some practical results are reviewed and
where: Rm = ––––– compared to the theoretical expected values predicted by
mc
mP/L is the mass of the passengers/load in kilograms, the TÜV method.
mC is the mass of the car in kilograms, The Canadian Study
aNL is the no load deceleration during a safety gear During the end of 1996, the Safety Engineering Services
application in mps2, Division of the Ministry of Municipal Affairs and Housing of
aFFFL is the deceleration during a safety gear applica- British Columbia carried out a study to look into the possi-
tion in free fall conditions in mps2, and bility of using the TÜV method to obviate the need to use
g is the acceleration due to gravity (9.81mps2). weights during the safety gear tests. The results have been
The interesting point to note is that the formula relies published in a report issued by TÜV [7]. Most of the tests
only on the ratio of the mass of passengers/load to the were carried out with an empty car, and only one test was
mass of the car, and not on their absolute values. carried out in the cases of both empty and fully loaded car.
Accepting the four assumptions given at the beginning of All the results for the cars with progressive safety gears
this section, we can see that the method gives a very simple have been extracted. The six lifts have been denoted with
Elevator Safeties and Governors
and practical tool to predict the free-fall full-load deceleration the arbitrary letters A to F. All lifts, except lift E, were only
under a safety gear application by just knowing the relative tested in the empty car condition. Lift E, had two tests:
mass of the car and full load, and making one measurement of One with an empty car and one with a fully loaded car.
acceleration during a safety gear application on an empty car. Nevertheless, for completeness, the results for all six lifts
Motivation and Benefit for Using the TüV Method have been tabulated in Table 1.
There are two main reasons for Table 1: Summary of Results from the British Columbia Study [7].
using the TÜV method, when doing Lift Rated Car Ratio Speed a_measured_ a_free fall a_measured
safety gear tests and when carrying out Load mass Rm empty calculated** full load
LG1 [6] inspections.
[kg] [mps] [mps2] [mps2] [mps2]
1. The first and most obvious reason is the
[kg]
possible damage that the full-load, full-
speed safety gear test could cause to the A 907 1813 0.50 1.25 8.76 2(2.57) N/A
lift components. It is known that such a B 680 1613 0.42 1.25 10.01 4(4.15) N/A
test does stress the main components in C 816 1804 0.45 1 7.88 2(2.39) N/A
the lift system, especially the gearbox.
D 1588 3270 0.49 1.5 18.91 9(9.47) N/A
2. The second main reason is the dif-
E 1011 2023 0.50 2 (1)* 13.2 5(5.53) 5.36
ference to timescales that the use of
weights has on a program of testing. F 1590 2900 0.55 4 (1.5)* 9.73 2(2.79) N/A
This is due to the difficulty of carrying and moving the *The no-load test was carried out at a reduced speed to ensure that the duration
weights, and the lack of storage space in some locations. was short enough to justify the assumption of slack rope.
**The value of a calculation by the ADIASYSTEM truncates the acceleration value to
This is not just an inconvenience, but it also presents a
a whole number. For completeness, I have added the exact value to two decimal
risk to the travelling public in delaying the LG1 tests, due places in brackets, based on Equation 6 shown earlier, and derived in Appendix C.
to a longer program. As mentioned earlier, only the ratio of the full load and
In addition to the motivation to use the TÜV method the car mass are needed in the formula for calculating the
and eliminate weight, there is actually an extra benefit predicted deceleration under free-fall full-load conditions
from using the method. It gives an extra result, which the (Equation 6 shown earlier and derived in Appendix C). So
full-load test does not give. It can predict what the free-fall the ratio of the two values is shown in a dedicated column.
full-load deceleration would be. This, in fact, is necessary A reduced speed test was carried out in the case of E and
to fully comply with the requirements of EN 81-1 clause F, in order to ensure that the counterweight and the traction
9.8.4 [8]: “For progressive safety gear, the average retardation effects could be cancelled.
in the case of free fall with rated load in the car shall lie The only comparison that could be done is the result of
between 0.2g and 1g.” case E, as that is the only one that had a no-load test in
addition to the full-load test. The numbers compare very 2. From that figure, the calculated free-fall full-load deceler-
well: 5.53 mps2 calculated, compared to 5.36 mps2 measured. ation value is (based on Equation 6 from Appendix C):
This shows that in this case, the effect of the counter- aNL – g x Rm 11.16 – 9.81 x 0.57
weight had been completely eliminated by the traction force aFFFL = ––––––––––– = ––––––––––––––––––––– = 3.54 mps2
1 + Rm 1.57
in the case of the full-load test. Examining the general case
Comparing this value with the value measured during
equation (derived in Appendix B, denoted Equation 1:
the full-load test, which is 6.62mps2, we note that the full-
FSG -FP/L -Fc +Fc/w -FT = aFL x(mP/L + mc ) . . .(1)
load test measured value is higher than the free-fall value.
As the result for this was identical to the free-fall case,
This points to the fact that the counterweight is assisting
we conclude that during a full-load safety gear applica-
the safety gear in retarding the fully loaded car, and that
tion: Fc/w = FT . In other words, the effect of the counter-
the counterweight effect has not been eliminated by the
weight is cancelled by the effect of the traction sheave force.
traction force from the sheave, as was the case of lift E in
This is obviously a special case, and it is not always the
the British Columbia project. In fact, the predicted figure by
case in all lifts that the force from the counterweight will be
the ADIASYSTEM is the worst case figure and is the figure
equal to the traction force from the sheave caused by the
needed to comply with the requirements of EN 81-1.
drive inertia.
Conclusions
Heavy-Duty Lift Results
The LG1 tests stipulate that a full-load rated speed test is car-
A no-load and full-load safety gear test was carried out
ried out on the safety gear, but allows the use of risk assess-
on a heavy-duty lift, during an LG1 test. The results and
ment to alter the tests or their frequency. There are concerns
parameters for the lifts and the test are tabulated Table 2.
regarding carrying out these tests for the
Table 2: Summary of Results from Heavy-duty Lift Test
cation is 118 milliseconds, calculated from these figures and Appendix A: Nomenclature
based on an actual speed of 1.32mps. The fact that the rope aNL deceleration during a safety gear application under
would have been slack validates one of the assumptions no loaded car with suspension intact [mps2]
required for the TÜV method, to allow the counterweight aFL deceleration during a safety gear application under
and traction forces to be ignored during the no-load test. full loaded car with suspension intact [mps2]
aFFF deceleration during a safety gear application of Using both of these points, and assuming that: aNL No
the car in full-load free-fall condition [mps2] load deceleration in mps2. The special case of the general
L equation (1), becomes:
g acceleration due to gravity [10mps2] FSG -Fc =aNL x (mc )
mC mass of the car [kg] FSG = aNL x (mc )+g x (mc )
mP/L mass of the load/passengers in the car [kg] FSG = (mc)x(aNL +g) (2)
mc/W mass of the counterweight [kg] Free Fall Case
Rm ratio of the full-rated passenger/load mass to the The second special case is when the suspension of the
mass of the car [ ] car fails completely, and the car goes into free fall under
FSG force of the safety gear [N] the acceleration of gravity. We will assume that the car is
FT traction force from the sheave on the ropes [N] fully loaded. Let us assume that: aFFFL Free-fall full-load
deceleration in mps2. The equations then become:
FC/W force due to the weight of the counterweight [N]
FSG -FP/L - Fc = aFFFL x (mc +mP/L )
FC force due to the weight of the car [N]
FSG =aFFFL x(mc +mP/L)+g x(mc +mP/L )
FP/L force due to the weight of the passengers/load in
FSG =( mc + mP/L )x(aFFFL +g) (3)
the car [N]
The importance of these three cases and their corre-
α counterbalance ratio [ ], where mc/w = mc + α x mP/L
sponding equations will become evident when deriving
Appendix B: Derivation of Safety Gear
the TÜV equation in the next section.
Application Equations
Appendix C: The TÜV Method Equation
In this Appendix, the equations for the general case of The calculation for the TÜV method will now be fur-
Elevator Safeties and Governors
safety gear application, and the equations for two special ther explored in order to derive the equation used. Using
cases are derived. assumptions 1 and 2, we see that the force from the
The General Case safety gear is the same at no load and full load free fall
The general case is derived first. This will then be used for conditions, coupled with the fact that we can ignore the
the two special cases. The general case assumes that none of effect of the rope during the no load safety gear applica-
the factors (discussed earlier) can be ignored, and that the sus- tion, gives us the following (from Figure 1):
pension is still intact (i.e., no suspension failure). Looking at a FSG = (mc + mP/L)x(aFFFL +g)= mc x (aNL +g)
free force diagram for the car and the passenger/load, we can ➞ aFFFL x(mc + mP/L)+g x mc +g x mP/L =g x mc + aNL x mc
derive the following (based on Figure 1). The total mass is: ➞ aFFFL x(mc + mP/L)= (aNL x mc)-(g x mP/L)
( )
mT =mP/L+mc (aNL x mc ) – (g x mP/L)
The resultant force on the car (assuming a positive ➞ aFFFL = ––––––––––––––––––––––– (4)
(mc + mP/L)
value represents an upward force):
Resultant Force=FSG -FP/L -Fc + Fc/w -FT Dividing throughout by mc, gives us the important result:
( (
)
)
Let us assume that the full-load deceleration under no mp/L
(aN/L) – g x –––––
load conditions is: aFL (full-load deceleration in mps2). We mc
aFFFL = –––––––––––––––––
then get the following result:
FSG -FP/L -Fc +Fc/w -FT = aFL x (mP/L +mc ) (1) ( mp/L
1 + –––––––
mc ) (5)