Force Isolation, Compensation For Mechanical Tolerances
Of Locks
By Ross Kinard
January 2009
The following article looks at a method, force isolation, to lower
the feedback given from mechanical locks reflecting the position of parts within
them.
Lock designs face many different problems with resistance to surreptitious
entry. One the major problems faced is feedback given off from the lock which
offers information on the placement of parts in the lock. If feedback is given
on the right parts, such as pins or discs operated on by a key, then it is possible
to determine the position those parts must be in to open. In design most locks
do not allow for feedback, but in the process of imperfect manufacturing, flaws
arise that do. These small flaws often caused by erosion of cutting materials
during manufacturing create parts in a lock which are slightly different sizes
and shapes. The degree of accuracy in the parts by design or manufacturer is
often called the tolerances of the lock. The higher the tolerances the more
similar the parts are and the less feedback is given off. Because manufacturers
will always have some flaws in their parts, their will always be some feedback
or change in the forces operating on them. Normally this feedback is given in
the form of resistance to opening.
This is where corrections can be made to compensate for a locks tolerance.
The degree of success made in surreptitious manipulation can only be measured
by the locks resistance to opening, but the locks resistance to opening does
not need to reflect pin or disc position (I will call these parts which must
be operated on by a key, or positioned correctly variable parts). So, this leads
to a need for a method to identify pin position with out transmitting the differences
in tolerances or resistance to the method for opening the lock. In a mechanical
lock, identifying variable part positions means making physical contact with
those parts. This physical contact means resistance and with resistance comes
opposing forces between those parts. This contact force represents positions
of variable parts, and will carry on to any parts connected to it with a contact
force running in any way the same direction. This contact force can not change
its own direction thou, and it can not affect a force running in a perpendicular
direction to itself. So, if the force used to open the lock runs in a perpendicular
direction to that of the force used to test variable parts, then they will not
reflect each other. This means the resistance to opening will no longer reflect
the position of variable parts.
Here are some examples of the problem.
In the pin tumbler to the right the contact force to test pin positions
is the same as the contact force to open the lock. The forces are in
the same direction or are parallel to each other. So the information
on the position of parts resulting from the contact force is completely
transferred. This means their is a complete reflection of the variable
parts position to the amount the lock will turn without opening.
In the combination lock the contact force to test variable parts runs
in a different direction then the contact force to open the lock. The
forces are supplied by the same spring, but that force is split in different
directions by the way the parts come in contact with each other. In
this case the force to open the lock represents a smaller portion of
the force to test variable parts. A portion of the force to test variable
parts is still represented thou. So, any changes in variable parts position
will still be transferred, and differences in positions can be identified.
|
|
 |
In this drawing you have a disc testing bolt
which comes in contact with the variable part or disc on a horizontal
plane. The locking bolt then hits the disc testing bolt on a vertical
plane. This directs the contact forces in perpendicular directions and
allows the parts to come in contact with each other with out allowing
their forces to affect each other.
In reality, manufacturing will not allow the locking bolt to hit the
disc testing bolt at a perfect 90 degree angle. The same problem of tolerances
will apply here and the two forces will interact on some level. Great
compensation can still be acquired thou. With a terrible 75 degree angle
cut on the end of the locking bolt there would be a 25.88% transfer of
contact force from the variable disc. A more realistic error of 5 degrees
in both the locking bolt and disc testing bolt would leave 17.37% transfer
of force. A hopefully realistic cut of 89 degrees in both the locking
bolt and disc testing bolt would give a 3.49% transfer of force. For grins,
an 89.9 degree cut in the locking bolt and disc testing bolt would leave
a 0.34% transfer of force.
So while perfect compensation still can not be acquired, feedback can
be greatly reduced by controlling the direction in which parts contact
each other. |
In this drawing you can see a more detailed
implementation. You have your variable discs in the back, two discs to
control the locking bars in the front, the locking bar or bolt on the
top, and an elevation bar used to hold the locking bar above the gates
of the variable discs, and a third bar or bolt which would be the dead
bolt. The variable discs would first be turned by a rotation of the key.
When they where set into place, as the key continued to rotate it would
turn the disc in the front allowing the locking bar on top to fall into
the gates of the variable discs. At the same time the dead bolt would
be pulled towards the locking bolt on top. If the locking bolt fell deep
enough into the gates, the dead bolt could be fully retracted, otherwise
it would be blocked by the locking bolt on top. When the key was rotated
back the dead bolt would be pushed back out and the locking bolt elevated
above the variable discs.
|
|
Quick notes for this design:
The distance between the dead bolt and locking bolt would have to be greater
then the distance between the highest point of the elevation bar and a distance
lower then the lowest possible gate.
Their would need to be some contact point to stop the locking bar as the
dead bolt came into contact with it. If it where allowed to just jam in the
variable disc gates slopes would being made, allowing more force from test
the variable discs to escape.
Other gears or cams could be connected to the variable discs to force and
attempt to manipulate them into a force pushing the dead bolt closed.
Some notes for future designs looking at this method for compensation:
Even with a perfect implementation of this method, it would be necessary
use a two stage method to prevent feedback. In a perfect
implementation the feedback does not come as a contact force, but a friction
force. The result of the contact from variable part being checked as it is moved.
The friction force could be made very small by lowering force applied to variable
parts, and keeping it separate from the force to open the lock, but it could
increased by moving the variable parts faster. So it could be measured. A two
stage method can perfectly prevent the feedback from the friction force
thou.
Gears, cams, wheels are all possible options to direct parts or forces in
the correct direction without forcing the end user to take multiple steps
in the unlocking process. A single force can be split safely in implementation.
Using gravity might also be an option to relocate parts. Care should be taken
in any method thou to prevent creating unwanted slopes in the lock, leaking
information of variable parts positions.
Slopes are not safe to be used in the lock design without worrying about
force directions even if they are only operating on a testing or unlocking
part at one time. For example, a slope used on a rotary disc to move the disc
testing bolt while not allowing itself or the disc testing bolt to come in
contact with the locking bolt would still give feedback. Even a slope acting
on the variable discs to move them from the disc testing bolt long enough
for it to be relocated could possibly give feedback.
This is a new idea for me, so I imagine there are more rule sets to be added
to this, as well as ones I may have been incorrect about. For any questions,
comments, or errors feel free to contact me at knowthebird@gmail.com.
gipoco.com
is neither affiliated with the authors of this page or responsible
for its contents. This is a safe-cache copy of the original web site.
gipoco.com
is neither affiliated with the authors of this page nor responsible
for its contents. This is a safe-cache copy of the original web site.